Allylic etherification via Ir(I)/Zn(II) bimetallic catalysis

Article abstract of DOI:10.1021/ol0508328

(Chemical Equation Presented) An efficient allylic etherification of aliphatic alcohols with allylic carbonates has been achieved by an iridium catalysis using stoichiometric zinc alkoxides or a two-component bimetallic catalytic system where the Ir(I) ca

Full text of DOI:10.1021/ol0508328

ORGANIC
LETTERS
2005
Vol. 7, No. 13
2679-2682
Allylic Etherification via Ir(I)/Zn(II)
Bimetallic Catalysis
Justin P. Roberts and Chulbom Lee*
Department of Chemistry, Princeton UniVersity, Princeton, New Jersey 08544
cblee@princeton.edu
Received April 15, 2005
ABSTRACT
An efficient allylic etherification of aliphatic alcohols with allylic carbonates has been achieved by an iridium catalysis using stoichiometric
zinc alkoxides or a two-component bimetallic catalytic system where the Ir(I) catalyst acts on allylic carbonates to generate electrophiles while
aliphatic alcohols are separately activated by Zn(II) coordination to function as nucleophilies. This reaction occurs with complete regiospecificity
and tolerates a wide range of functional groups.
Development of catalysts that mobilize otherwise inert
substrates to engage in chemical reactions continues to be a
major theme in organic chemistry. While the majority of
catalytic reactions are typically mediated by a single catalyst,
opportunities for further development arise from employing
a two-component catalyst system in which the nucleophilic
and electrophilic components of the reactant are activated
independently by respective catalysts.1 Such approaches,
though uncommon, have demonstrated high potential for the
discovery of new reactions, particularly in the processes
involving a η3-allylmetal species.2
Recently, the scope of the transition metal-catalyzed allylic
alkylation has been extended to include the allylic etherifi-
cation of aliphatic alcohols by using copper(I)3 and
zinc(II)4 alkoxides as nucleophiles. These discoveries have
stimulated the emergence of powerful strategies for the
synthesis of ether linkages conjoining chiral carbons, as
exemplified by the stereodivergent approach to oxacycle5 and
glycosidic bond formations.4b While investigating the pal-
ladium-catalyzed allylic etherification,4 we wondered whether
the zinc effect might also be operative in the context of other
transition metal catalysts. In particular, it was of interest to
examine these catalyst systems which could give different
stereo- and regiochemical outcomes vis-a`-vis the Pd(0)/
Zn(II) system, thus broadening the utility of the zinc alkoxide
protocol.6 Moreover, the facile ligand exchange process of
zinc(II) complexes was envisaged to be exploited for the
catalytic activation of the alcohol,7 whereby the etherification
could be achieved through catalysis mediated by two separate
transition metal complexes. This approach would provide a
unique opportunity to modulate the reactivity of each reactant
and/or to achieve asymmetric induction by independent
tuning of each metal coordination sphere. We report here
an iridium-catalyzed allylic etherification of aliphatic alcohols
that proceeds with high regiospecificity in the presence of a
stoichiometric or catalytic zinc alkoxide.
(1) For reviews on bimetallic and bifunctional catalyses, see: (a) van
Den Beuken, E. K.; Feringa, B. L. Tetrahedron 1998, 54, 12985. (b)
Rowlands, G. J. Tetrahedron 2001, 57, 1865.
(2) (a) Sawamura, M.; Sudoh, M.; Ito, Y. J. Am. Chem. Soc. 1996, 118,
3309. (b) Trost, B. M.; McEachern, E. J.; Toste, F. D. J. Am. Chem. Soc.
1998, 120, 12702. (c) Kamijo, S.; Yamamoto, Y. Angew. Chem., Int. Ed.
2002, 41, 3230. (d) Kamijo, S.; Yamamoto, Y. J. Org. Chem. 2003, 68,
4764. (e) Jellerichs, B. G.; Kong, J.-R.; Krische, M. J. J. Am. Chem. Soc.
2003, 125, 7758.
To assess the feasibility of the proposed bimetallic catalytic
process, the etherification was first examined with allylic
carbonate 1a using a stoichiometric zinc benzyloxide and a
(5) (a) Evans, P. A.; Leahy, D. K.; Andrews, W. J.; Uraguchi, D. Angew.
Chem., Int. Ed. 2004, 43, 4788. (b) Shu, C.; Hartwig, J. F. Angew. Chem.,
Int. Ed. 2004, 43, 4794.
(3) (a) Evans, P. A.; Leahy, D. K. J. Am. Chem. Soc. 2002, 124, 7882.
(b) Evans, P. A.; Leahy, D. K.; Slieker, L. M. Tetrahedron: Asymmetry
2003, 14, 3613.
(4) (a) Kim, H.; Lee, C. Org. Lett. 2002, 4, 4369. (b) Kim, H.; Men, H.;
Lee, C. J. Am. Chem. Soc. 2004, 126, 5, 1336.
(6) For the compatibility of zinc alkoxides with iridium catalyses, see:
(a) Miyabe, H.; Matsumura, A.; Moriyama, K.; Takemoto, Y. Org. Lett.
2004, 6, 4631. (b) Miyabe, H.; Yoshida, K.; Yamauchi, M.; Takemot, Y.
J. Org. Chem. 2005, 70, 2148. (c) Ref 5b.
(7) Parkin, G. Chem. Commun. 2000, 1971.
10.1021/ol0508328 CCC: $30.25
© 2005 American Chemical Society
Published on Web 06/01/2005

Table 1. Iridium-Catalyzed Allylic Etherification of Alcohol 2a with Carbonate 1a Using a Stoichiometric Zinc Reagenta
entry
catalystb
[Ir(COD)Cl]2
ligandc
salt
time
yieldd
3aa:4aae
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16f
P(OPh)3
P(OAr)3
pybox
NH4I
NH4I
NH4I
NH4I
NH4I
NH4I
NH4I
NH4I
48 h
2 h
39%
90%
74%
96%
0%
0%
4%
46%
0%
0%
50%
80%
14%
11%
95%
40%
2:3
22:1
>50:1
>50:1
[Ir(COD)Cl]2
[Ir(COD)Cl]2
[Ir(COD)Cl]2
[Ir(COE)2Cl]2
Ir(CO)(dppe)Cl
Ir(CO)(PPh3)2Cl
Ir(PCy3)(COD)(Py)PF6
[Ir(COD)Cl]2
[Ir(COD)Cl]2
[Ir(COD)Cl]2
[Ir(COD)Cl]2
[Ir(COD)Cl]2
[Ir(COD)Cl]2
[Ir(COD)Cl]2
Pd(PPh3)4
0.25 h
0.25 h
48 h
48 h
48 h
48 h
48 h
48 h
48 h
12 h
48 h
48 h
24 h
1.5 h
pyridine
5:1
>50:1
pyridine
pyridine
pyridine
pyridine
pyridine
pyridine
pyridine
NH4OAc
NH4BF4
NH4F
NH4Cl
NH4Br
NMe4I
2:1
1:1
1:2
15:1
1:2
<1:50
a All reactions were carried out in THF (1.0 M) with 1.0 equiv of 1a and 1.1 equiv of 2a at 25 °C in the presence of 5 mol % [M]/ligand, 0.55 equiv of
Et2Zn, and 1.0 equiv of salt. b COD ) 1,5-cyclooctadiene; COE ) cyclooctene. c P(OAr)3 ) tris(2,4-di-tert-butylphenyl)phosphite; pybox ) 2,6-bis[(4R)-
1
(+)-isopropyl-2-oxazolin-2-yl]pyridine; DTBBP ) di(tert-butyl)-2-biphenylphosphine. d Isolated yields. e Determined by H NMR. f Ref 4.
variety of iridium catalysts (Table 1). Guided by literature
precedents in which the Ir-catalyzed allylic alkylation reac-
tions with a substrate of type 1a generated a branched product
such as 3aa in preference to the linear isomer 4aa, our studies
were focused on the identification of catalyst systems viable
with a zinc alkoxide and analysis of their regiochemical
outcome.8,9 Initial experiments performed in THF using 2.5
mol % [Ir(COD)Cl]2 and 5 mol % P(OPh)3 resulted in a poor
yield of allylic ethers 3aa and 4aa as an inseparable 1:1.5
isomeric mixture (entry 1). Notably, a considerable amount
of benzaldehyde (ca. 10%) was formed presumably through
a â-H elimination of an iridium benzyloxide. It was thus
reasoned that the addition of more steric bulk to the ligand
would minimize such a process occurring at the iridium
center. Accordingly, the employment of a tert-butyl-
substituted ligand substantially increased both the yield and
the regioselectivity (entry 2). Further improvement came
surprisingly from the use of nitrogen ligands.10 In the
presence of pybox and pyridine, the reactions were complete
in 15 min to furnish 3aa as a single product, whereas no
reaction occurred without these ligands (entries 3 and 4).
The screening of various precatalysts, however, proved to
be less fruitful, as COE- and CO-bound Ir complexes per-
formed poorly (entries 5-8). Only Crabtree’s catalyst
exhibited modest results, highlighting the critical role of 1,5-
cyclooctadiene (COD) as a requisite bidentate ligand in this
catalyst system (entry 8).11 Also noteworthy were the
significant salt effects, which proved to be most beneficial
when NH4I was used (entry 4 vs entries 9-15).12 Both the
ammonium and iodide components of NH4I were required
for high turnover and regioselectivity. The regiochemical
outcome could be reoriented by using Pd(PPh3)4, which led
to the exclusive formation of the linear isomer 4aa (entry
16).4a
Having established an optimized set of conditions for the
Ir (catalytic)-Zn (stoichiometric) protocol, we then set out
to screen ligands for the zinc center and probed the prospect
of achieving bimetallic catalysis (Table 2). In control
experiments performed without an additional ligand, the yield
of the allylation products was roughly proportional to the
amount of Et2Zn employed. Hence, attempts were first made
using common zinc(II) ion ligands derived from sp2-
hybridized nitrogens (entries 1-4).13 These ligands indeed
promoted turnover of the zinc catalyst but led to prolonged
(8) (a) Takeuchi, R.; Kashio, M. Angew. Chem., Int. Ed. Engl. 1997,
36, 263. (b) Takeuchi, R.; Kashio, M. J. Am. Chem. Soc. 1998, 120, 8647.
(c) Janssen, J. P.; Helmchen, G. Tetrahedron Lett. 1997, 38, 8025.
(9) For recent examples of carbon-heteroatom bond formation by Ir-
catalyzed allylic alkylations, see: (a) Takeuchi, R.; Ue, N.; Tanabe, K.;
Yamashita, K.; Shiga, N. J. Am. Chem. Soc. 2001, 123, 9525. (b) Ohmura,
T.; Hartwig, J. F. J. Am. Chem. Soc. 2002, 124, 15164. (c) Lopez, F.;
Ohmura, T.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125, 3426. (d) Kiener,
C. A.; Shu, C.; Incarvito, C.; Hartwig, J. F. J. Am. Chem. Soc. 2003, 125,
14272. (e) Miyabe, H.; Yoshida, K.; Matsumura, A.; Yamauchi, M.;
Takemoto, Y. Synlett 2003, 567. (f) Miyabe, H.; Matsumura, A.; Yoshida,
K.; Yamauchi, M.; Takemoto, Y. Synlett 2003, 2123. (g) Lipowski, G.;
Helmchen, G. Chem. Commun. 2004, 116. (h) Welter, C.; Kock, O.;
Liposwky, G.; Helmchen, G. Chem. Commun. 2004, 896. (i) Fischer, C.;
Defieber, C.; Suzuki, T.; Carreira, E. M. J. Am. Chem. Soc. 2004, 126,
1628. (j) Miyabe, H.; Matsumura, A.; Moriyama, K.; Takemoto, Y. Org.
Lett. 2004, 6, 4631.
(10) For the use of a pybox-type ligand in the Ir-catalyzed allylic
alkylation, see: (a) Lavastre, O.; Morken, J. P. Angew. Chem., Int. Ed.
1999, 38, 3163. (b) Ref 9j.
(11) (a) Glorius, F. Angew. Chem., Int. Ed. 2004, 43, 3364. (b) For an
X-ray crystal structure of a COD-bound η3-allyliridium complex, see:
Bartels, B.; Garcia-Yebra, C.; Rominger, F.; Helmchen, G. Eur. J. Inorg.
Chem. 2002, 2569. (c) For a chiral dialkene ligand in an Ir-catalyzed
reaction, see ref 9i.
(12) For a review, see: (a) Fagnou, K.; Lautens, M. Angew. Chem., Int.
Ed. 2002, 41, 1, 26. (b) For an example of profound halide effects on the
regioselectivity in a Rh-catalyzed allylic etherification, see ref 3.
2680
Org. Lett., Vol. 7, No. 13, 2005

Table 2. Screening of Ligands for the Zn(II) Center in the Ir(I)/Zn(II)-Catalyzed Allylic Etherificationa
entry
ligandb
time
yieldc
3aa:4aad
entry
ligandb
time
yieldc
3aa:4aad
1
2
3
4
5
6e
7
8
BIPY
TRPY
Tpm
KTp
NME
NME
Gly
48 h
48 h
48 h
48 h
0.5 h
0.5 h
0.75 h
0.5 h
8%
36%
59%
29%
65%
80%
67%
74%
11:1
22:1
47:1
>50:1
34:1
>50:1
>50:1
>50:1
9
10
11e
12
12
14
15
L-His
L-Trp
L-Trp
L-Asn
L-Tyr
L-Ser
L-Cys
0.5 h
0.25 h
0.5 h
0.75 h
0.25 h
0.5 h
73%
89%
70%
68%
68%
70%
68%
>50:1
>50:1
>50:1
>50:1
>50:1
>50:1
>50:1
0.5 h
L-Glu
a All reactions were performed with 1.0 equiv of 1a and 1.5 equiv of 2a in THF (1.0 M) at 25 °C unless otherwise noted. b BIPY ) 2,2′-bipyridine; TRPY
) 2,2′-6′,2′′-terpyridine; Tpm ) tris(pyrazol-1-yl)methane; KTp ) potassium hydrotris(pyrazol-1-yl)borate; NME ) (1R,2S)-(-)-N-methylephedrine. c Isolated
1
yields. d Determined by H NMR. e Reactions were performed in the absence of pyridine.
reaction times. In contrast, marked acceleration of the
reaction rate was brought about when a â-amino alcohol
(NME)14 (entries 5 and 6) and various R-amino acids (entries
7-15) were employed as ligands. In these cases, all reactions
proceeded rapidly with complete regioselectivity in the
presence of these N,O-chelators, among which L-tryptophan
displayed superior results (entries 10 and 11). An additional
finding was that pyridine was not required when the reactions
were carried out with these ligands (entries 6 and 10),
although the use of pyridine led to a shorter reaction time
and a higher yield (entry 10 vs entry 11).
With the two iridium-catalyzed protocols viable under
stoichiometric or catalytic zinc conditions, we examined the
scope of the new processes. As shown in Table 3, a variety
Table 3. Iridium-Catalyzed Allylic Etherification Using Stoichiometric and Catalytic Zinc Alkoxidesa
a Method A: 1.0 equiv of 1; 1.1 equiv of 2; 2.5 mol % [Ir(COD)Cl]2; 5 mol % pyridine; 0.55 equiv of Et2Zn; 1.0 equiv of NH4I; 25 °C; THF (1.0 M).
Method B: 1.0 equiv of 1; 1.5 equiv of 2; 2.5 mol % [Ir(COD)Cl]2; 5 mol % pyridine; 10 mol % Et2Zn; 10 mol % L-Trp; 10 mol % NH4I; 25 °C; THF
(1.0 M). b First and second numbers refer to the isolated yields from methods A and B, respectively. c Diastereomeric mixture (1:1). d Single isomer.
Org. Lett., Vol. 7, No. 13, 2005
2681

of allylic ethers could be prepared with complete regiocontrol
from the reactions of structurally diverse aliphatic alcohols
and allylic carbonates. Notable from these examples was the
high chemoselectivity manifested through exceptional toler-
ance of a wide range of functional groups. Both mono- and
disubstituted carbonates participated well in the etherification
process under the standard conditions using stoichiometric
zinc reagents (conditions A), whereas the catalytic zinc
conditions (conditions B) failed to induce a reaction when
applied to disubstituted allylic substrates 1e-g. It was worthy
of note, however, that the catalytic zinc conditions proved
to be more effective in many cases with monosubstituted
allylic substrates. Reaction rates of the stoichiometric and
catalytic methods were comparable, with slightly longer
reaction times for the latter (5-20 vs 15-60 min). As
demonstrated by the formation of single isomers as the
products, the reactions of disubstituted allylic substrates
proceeded with complete retention of regio- and stereochem-
istry (entries 11-13). In particular, a high level of selectivity
was maintained in the reaction of dimethylallyl system 1e,
in which the scrambling of regiochemistry via a symmetric
irido-π-allyl intermediate would constitute loss of stereo-
chemistry (entry 11). In contrast, a disparity between the
regio- and stereochemical outcomes was noted in reactions
of monosubstituted allylic systems. For example, when
optically enriched (S)-1b (98% ee) was reacted with 2d under
conditions A, ether 3bd was produced as a single regioisomer
but only in 57% ee (cf. entry 3). These results indicated that
with the present catalytic system, the putative “enyl” iridium
intermediate undergoes slow allylic isomerization but fast
enantiofacial interconversion processes relative to the inter-
molecular nucleophilic addition reaction with a zinc alkox-
ide.15,16 The regiospecific nature of the present Ir protocol
was also manifested in the reaction of the linear substrate 7
(Scheme 1), whereas the Pd-catalyzed method proved to be
regioselective (cf. entry 16, Table 1).
Scheme 2. Proposed Mechanism for the Ir(I)/Zn(II) Bimetallic
Catalytic Allylic Etherification
catalytically and independently by the respective iridium(I)
and zinc(II) catalysts. Under both stoichiometric and catalytic
zinc conditions, the iridium catalyst acts on the allylic
carbonate to generate the requisite cationic η3-allyl iridium
intermediate (left catalytic cycle). Critical to the bimetallic
catalytic mechanism is an alkoxide exchange at the zinc
center (right catalytic cycle). In light of the well-known pKa-
dependent equilibrium between a free alcohol and its
corresponding zinc alkoxide,7 it is believed that alcohols with
a higher acidity than tert-butyl alcohol are capable of
undergoing the exchange and subsequent nucleophilic ad-
dition processes. An additional important feature that renders
the dual catalytic cycles feasible is the poor reactivity of
zinc tert-butoxide toward the iridium-bound cationic inter-
mediate, whereby the requisite deprotonation process is
effectively mediated by zinc tert-butoxide. While a detailed
understanding awaits further studies, the pronounced effect
of NH4I may be due to the combination of the high affinity
of the iodide anion for the iridium center and a subtle proton-
transfer process required for the reaction.17
In summary, we have demonstrated that “softening-the-
alkoxide approach” via zinc(II) coordination is viable for
the regiospecific O-allylation of aliphatic alcohols under
iridium(I) catalysis. High chemo- and regioselectivities and
operational simplicity are important features of the new
protocol. Our studies have also led to the development of a
two-component catalyst system that achieves the allylic
etherification through the separate yet cooperative action of
the iridium and zinc complexes on the respective allylic
carbonate and alcohol substrates. On the basis of the present
findings, future investigations will be directed at the develop-
ment of related catalytic reactions and their asymmetric
variants, particularly by crafting the ligands independently
for each metal center.
Scheme 1. Regiospecificity of Allylic Etherifications
Acknowledgment. We thank Princeton University for
generous support of this research. We would also like to
thank Merck for the donation of compound 2m.
The mechanism for the Ir/Zn two-component bimetallic
catalytic allylic etherification is proposed in Scheme 2, in
which the allylic carbonate and aliphatic alcohol are activated
Supporting Information Available: Experimental pro-
cedures and spectral data. This material is available free of
charge via the Internet at http://pubs.acs.org.
(13) Parkin, G. Chem. ReV. 2004, 104, 699.
(14) (a) Frantz, D. E.; Fa¨ssler, R.; Carreira, E. M. J. Am. Chem. Soc.
2000, 122, 1806. (b) Frantz, D. E.; Fassler, R.; Tomooka, C. S.; Carreira,
E. M. Acc. Chem. Res. 2000, 33, 373.
(15) Evans, P. A.; Nelson, J. D. J. Am. Chem. Soc. 1998, 120, 5581.
(16) (a) For discussion on an irido-enyl species, see ref 11b. (b) The
rapid enantiofacial interconversion of an irido-η3-allyl intermediate was
utilized for asymmetric induction via deracemization. See: (b) Bartels, B.;
Helmchen, G. Chem. Commun. 1999, 741.
OL0508328
(17) (a) Tobe, M. L. In ComprehensiVe Coordination Chemistry;
Wilkinson, G., Gillard, R. D., McCleverty, J. A., Eds.; Pergamon: New
York, 1987; Vol. 1, pp 281-329. (b) Also see ref 12a.
2682
Org. Lett., Vol. 7, No. 13, 2005