Achieving chemo-, regio-, and stereoselectivity in palladium-catalyzed reaction of γ-borylated allylic acetates

Article abstract of DOI:10.1021/ol201661s

Three-carbon highly functionalized γ-borylated allylic acetates underwent a regio- and stereocontrolled Tsuji-Trost reaction in the presence of palladium complexes. An ipso substitution of the acetate with complete stereoretention of the chiral center was achieved, leading to vinylic boronates with enantiomeric excesses above 99%.

Full text of DOI:10.1021/ol201661s

ORGANIC
LETTERS
2011
Vol. 13, No. 15
4132–4135
Achieving Chemo-, Regio-, and
Stereoselectivity in Palladium-Catalyzed
Reaction of γ-Borylated Allylic Acetates
Krishna Kishore Kukkadapu,† Aziz Ouach,† Pedro Lozano,‡ Michel Vaultier,†,§ and
Mathieu Pucheault*,†,§
ꢀ
ꢀ
Chimie et Photonique Moleculaires, UMR CNRS 6510, Universite Rennes 1, Campus de
Beaulieu, 263 Avenue du General Leclerc, F-35042 Rennes Cedex, France,
ꢀ ꢀ
´
Departamento de Bioquımica y Biologıa Molecular B e Inmunologıa, Facultad de
Quiımica, Universidad de Murcia, P.O. Box 4021, E-30100 Murcia, Spain, and Institut
´
ꢀ
des Sciences Moleculaires, UMR CNRS No. 5255, Universite Bordeaux 1, 351 Cours de
la Liberation, F-33405 Talence Cedex, France
ꢀ
ꢀ
m.pucheault@ism.u-bordeaux1.fr
Received June 22, 2011
ABSTRACT
Three-carbon highly functionalized γ-borylated allylic acetates underwent a regio- and stereocontrolled TsujiꢀTrost reaction in the presence of
palladium complexes. An ipso substitution of the acetate with complete stereoretention of the chiral center was achieved, leading to vinylic
boronates with enantiomeric excesses above 99%.
Organometallic catalysis has arisen as a special key tool
in modern synthetic methods, especially for pharmaceuti-
cal preparation. In many cases, the flexibility of transition
metal complexes offers unusual reactivity and complemen-
tary methods for synthesis. Indeed, many transformations
are catalyzed by palladium complexes, including Heck
reactions, SuzukiꢀMiyaura and Sonogashira cross cou-
pling, TsujiꢀTrost allylic substitution, and Hartwigꢀ
Buchwald N-arylations.1 However, despite a plethora of
publications reporting the development of new catalysts
for promoting such transformations, the issue of chemos-
electivity is usually poorly addressed when substrates bear
multiple putative reaction sites.
challenges of engendering regioisomers and displaying a
chiral center on both the starting material and the products.
The preparation of these compounds is quite straightfor-
wardstarting fromthe corresponding propargylic acetates.
A one-pot procedure has been optimized, consisting of a
hydroboration using diisopinocamphenylborane followed
by a refunctionalization at the boron atom using acetalde-
hyde. The resulting diethyl boronates were transesterified
in situ by pinacol leading to the stable and easy to handle
vinylpinacol boronates. Such compounds can alternatively
be prepared following other reported methods. Hydrobora-
tion using dicyclohexylborane followed by oxidation with
trimethylamine oxide leads to similar yields.2 Direct hydro-
boration with pinacolborane of the TMS protected pro-
pargylic alcohol followed by deprotection with citric acid
and subsequent acetylation3 turned out equally effective
but slightly longer (Scheme 1).
We sought to contribute to this issue by using γ-borylated
allylic acetates as coupling partners. These highly functio-
nalized three-carbon building blocks offer the additional
†
ꢀ
Universite Rennes 1.
‡ Universidad de Murcia.
§
ꢀ
Universite Bordeaux 1.
(2) Pietruszka, J.; Witt, A. J. Chem. Soc., Perkin Trans. 1 2000, 4293–
4300.
(3) Fortineau, A.-D.; Robert, M.; Gueguan, J.-P.; Carrie, D.; Mortier,
J.; Vaultier, M. C. R. Acad. Sci. Serie IIc 1998, 1, 253–257.
(1) Applied Homogeneous Catalysis with Organometallic Compounds:
A Comprehensive Handbook in Three Volumes; Cornils, B., Herrmann,
W. A., Eds.; Wiley-VCH: Weinheim, 2002; Vols. 1ꢀ3.
r
10.1021/ol201661s
Published on Web 07/06/2011
2011 American Chemical Society

complexes.10 However, Walsh recently showed that, on
similar substrates bearing a boron in the β position, a
selective nucleophilic substitution could be achieved using
carbon- and nitrogen-based nucleophiles.11
Scheme 1. Chemo-, Regio-, and Stereoselectivity Issues
We present here our recent studies related to the chemo-,
regio-, and nucleophilic substitution on γ-borylated allylic
acetates. After optimization of reactions conditions, it
turned out that 1% Pd(OAc)2/3% PPh3 in THF was a
general catalytic system, leading to selective ipso sub-
stitution (Table 1, entry 1).
Table 1. Catalytic System Optimization
Considering the mechanistic difference between oxida-
tive addition to the carbonꢀoxygen bond leading to a
π-allyl complex and the transmetalation from boron to
palladium, we first aimed for a selective allylic substitution.4
To our knowledge, only a few examples of allylic susbtitu-
tion on similar substrates are known. Asymmetric allylic
substitution using a Grignard reagent can be catalyzed by
phosphoramiditeꢀcopper complexes, leading to exclusive
formation of the SN20 products with excellent regio- and
stereocontrol.5 Similarly, iridiumꢀphosphoramidite com-
plexes react on these allylic acetates, leading to the product
substitutedR to boron withgood regio-andstereocontrol.6
However, in those cases reported by Hall, chemoselectivity
is a lower concern as transmetalation of boron to iridium
or copper is relatively sluggish although theoretically
possible. In the presence of palladium catalysts, few examples
have been reported, but they confirm that good chemos-
electivitycanbeachievedwhenusinghighlyreactivespecies.
For instance, diazomethane reacts smoothly with γ-borylated
allylic acetates to form the corresponding cyclopropyl-
boronates.7 Favoring the SuzukiꢀMiyaura cross coupling
is also possible8 when tert-butyl carbonate and activated
aryl iodides are used or when water is used as a solvent.9
However, achieving selective palladium-catalyzed allylic
substitution with mild nucleophiles is a much greater
challenge and leads “to mixtures of regioisomeric deboro-
nation products and other unidentified materials” as Hall
noticed. Indeed, some cross-coupling between allylic acet-
ates and boronic acids has already been described, showing
the wide range of reactions promoted by palladium
entry
[Pd]
Pd(OAc)2
ligand (%)
yielda (%)
1
2
3
4
5
6
7
8
PPh3 (3)
PPh3 (3)
PPh3 (3)
77
72
76
0
PdCl2
[Pd(allyl)Cl]2
Pd(OAc)2
Pd(OAc)2
PPh3 (4)
PPh3 (2)
PPh3 (2)
77
70
70
75
Pd2(dba)3-CHCl3
Pd(dba)2
Pd(PPh3)4
a Isolated yield after purification by flash chromatography.
Using [Pd(allyl)Cl]2 (Table 1, entry 3) or PdCl2 (Table 1,
entry 2) led to similar activities, albeit somehow slightly
less efficient. Phosphine addition was required for good
conversion (Table 1, entries 4 and 5) and more than three
equivalents did not improve the system (Table 1, entry 5).
Other palladium sources were less efficient, especially
when introduced under a Pd(0) oxidation state as in Pd2-
(dba)3ꢀCHCl3 or Pd(dba)2 (entries 6 and 7). Only Pd-
(PPh3)4 allowed yields similar to those obtained when
Pd(II) was reduced in situ. Among ligands, other phosphines
were at least as efficient as triphenylphosphine. NHC-
carbenes were unsuccessful, and nitrogen-based ligands
failed. Overall, given its simplicity and cost, triphenylpho-
sphine was kept in the optimized catalytic system. Among
other solvents, only THF led reproducibly to good conversions.
These mild conditions allowed us to perform Tsujiꢀ
Trost reactions with various nucleophiles. We began our
studies using enolate-type nucleophiles derived from
1,3-dicarbonyl compounds. Alkyl malonates were efficient
regardless of the ester substitution (Table 2, entries 1, 5,
and 9); β-ketoesters (Table 2, entries 2, 6, and 10), 1,
3-diketones (Table 2, entries 4 and 8), and cyanoacetate
(Table 2, entries 3, 7, and 11) reacted equally well. Even
when using a sterically hindered nucleophile such as 2e,
quaternary centers were obtained in similar yields (Table 2,
entries 13 and 14). In the case of 1d, the only isolated
(4) (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395–
422. (b) Trost, B. M.; Crawley, M. L. Chem. Rev. 2003, 103, 2921–2944.
(5) Carosi, L.; Hall, D. G. Angew. Chem., Int. Ed. 2007, 46, 5913–
5915.
(6) Peng, F.; Hall, D. G. Tetrahedron Lett. 2007, 48, 3305–3309.
(7) (a) Hohn, E.; Pietruszka, J.; Solduga, G. Synlett 2006, 1531–1534.
(b) Pietruszka, J.; Solduga, G. Eur. J. Org. Chem. 2009, 5998–6008.
ꢁ
ꢀ
ꢁ
ꢀ
(8) Stambasky, J.; Malkov, A. V.; Kocovsky, P. Collect. Czech.
Chem. Commun. 2008, 73, 705–732.
(9) Genet, J. P.; Linquist, A.; Blart, E.; Mouries, V.; Savignac, M.;
Vaultier, M. Tetrahedron Lett. 1995, 36, 1443–1446.
(10) (a) Ohmiya, H.; Makida, Y.; Tanaka, T.; Sawamura, M. J. Am.
Chem. Soc. 2008, 130, 17276–17277. (b) Su, Y.; Jiao, N. Org. Lett. 2009,
11, 2980–2983. (c) Comer, E.; Organ, M. G.; Hynes, S. J. J. Am. Chem.
Soc. 2004, 126, 16087–16092.
(11) Hussain, M. M.; Walsh, P. J. Angew. Chem., Int. Ed. 2010, 49,
1834–1837.
Org. Lett., Vol. 13, No. 15, 2011
4133

entry 5). Modifying the diamine scaffold led to a dramatic
ee decrease (30%, Table 3, entry 4) but favored the other
enantiomer.
Table 2. TsujiꢀTrost Reaction on γ-Borylated Allylic Acetates
Scheme 2. One-Pot TsujiꢀTrost SuzukiꢀMiyaura Reaction
entry
acetate
Nu
product
yield (%)
1
2
1a
1a
1a
1a
1b
1b
1b
1b
1c
1c
1c
1d
1a
1c
2a
2b
2c
2d
2a
2b
2c
2d
2a
2b
2c
2a
2e
2e
3aa
3ab
3ac
3ad
3ba
3bb
3bc
3bd
3ca
3cb
3cc
3da
3ae
3ce
77
80
80
80
80
83
79
82
74
76
77
61
75
79
3
4
5
6
7
8
9
10
11
12
13
14
To our delight, performing the allylic substitution on a
chiral substrate (S)-1a (ee >99%) led to 3aa with 88% ee,
representing a mere 5% racemization of the substrate
(Table 3, entry 6). The size of this memory effect is some-
how unprecedented, especially considering the 4 h reaction
time, giving to the generated π-allyl complex more than
enough time to racemize. Several examples could be found
in the literature trying to explain this memory effect.
According to Pfaltz and Helmchen,12 regioselectivity of
the reaction is directly related to SN1 vs SN2 type mechan-
ism, the latter favoring the linear product. In our case, the
complete ipso substitution corroborates a double SN2
mechanism. This was confirmed by the absolute config-
uration of the carbon center, which appeared to be formed
with a complete stereoretention.13 According to several
reports,14 the addition of LiCl often improves this stereo-
retention effect. In our hands, it led to complete inactiva-
tion of the catalytic system. Overall, this reaction is supposed
to proceed through a true π-allyl palladium complex
bearing a boron subsituent.15
a Isolated yield after purification by flash chromatography.
boron-containing product is the unconjugated ipso-
substituted product, which is unexpected considering the
traditional outcome of the TsujiꢀTrost reaction in the
presence of palladium complexes. However, another pro-
duct is isolated in 15% yield, which could originate either
from direct coupling of the boronate moiety with the
malonate or from the SN20 product protodeboronation
during workup. This emphasizes nicely the positive influ-
ence of the boron substituent on the regioselectivity of the
reaction (Table 2, entry 12, and Scheme 2).
Based on our success in performing selective Tsujiꢀ
Trost substitutions, wethenpursuedour study by usingthe
obtained vinyl boronate in a SuzukiꢀMiyaura cross cou-
pling. Compound 3aa led to the styrene derivative in 91%
yield in the presence of phenyl iodide and Pd(OAc)2. In the
one-pot sequence beginning with allylic substitution, the
residual palladium species efficiently catalyzed the second
cross-coupling reaction. We ultimately obtained the
double cross-coupled adduct in a slightly better yield than
by performing the two steps separately.
In the continuity of controlling the regio- and chemos-
electivity of the reaction, we then tackled the stereocontrol
issue. Overall, starting from racemic allylic acetate, using
DACH-phenyl Trost ligand L1 in lieu of triphenylpho-
sphine, a 78% ee was observed (Table 3, entry 2). Other
chiral ligands from the same family were tested. The
simplest ligand turned out to give the best selectivities, as
the naphthyl-derived ligand L2 led to 54% ee (Table 3,
entry 3) and the pyridyl ligand L4 led to 11% ee (Table 3,
Combining in a matched fashion, (S,S)-DACH-phenyl
Trost ligand and chiral (S)-starting material afforded the
boronate with >99% ee (Table 3, entry 8). More surpris-
ing was the 97% ee of the same enantiomer obtained
with the mismatched pair (R,R)-ligand and (S)-boronate,
(12) Helmchen, G; Pfaltz, A. Acc. Chem. Res. 2000, 336–345.
(13) This stereochemical assignment was attributed by comparison
with the SuzukiꢀMiyaura cross-coupled product 4aa, of which the
optical rotation was already reported.
(14) (a) Amatore, C.; Jutand, A.; M0Barki, M. A.; Meyer, G.;
Mottier, L. Eur. J. Inorg. Chem. 2001, 873–888. (b) C. Lloyd-Jones,
G.; C. Stephen, S. Chem. Commun. 1998, 2321–2322.
(15) Burks, H. E.; Kliman, L. T.; Morken, J. P. J. Am. Chem. Soc.
2009, 131, 9134–9135.
4134
Org. Lett., Vol. 13, No. 15, 2011

entry 9). Using this three-carbon template bearing a boron
substituent provides an interesting alternative to the direct
TsujiꢀTrost reaction owing to the versatility of boron
chemistry. For example, in the case of substrate bearing
similar aryl groups on both ends of the π-allyl palladium
intermediate, it is hard, albeit possible, to control nucleo-
philic attack.16 The one-pot procedure provides an inter-
esting alternative for controlling this regioselectivity. The
tandem TsujiꢀTrost/SuzukiꢀMiyaura cross coupling be-
tween 4-methylphenyl iodide, phenyl-substituted allylic
acetate, and dimethyl malonate afforded40dain 78% yield.
The one-pot sequence was superior to the stepwise process
in improving the overall yield to 78% instead of 45%
(Scheme 2). Additionally, if required, the boron substitu-
ent could be replaced by a large number of functional
groups using more classical transformations such the
ChanꢀLamꢀEvans coupling, including the use of sodium
azide as a nitrogen counterpart,17 leading to product 5aa in
78% yield (Scheme 3).
Table 3. TsujiꢀTrost Reaction on γ-Borylated Allylic Acetates
Overall, we have managed to use a highly functionalized
three-carbon building block in a chemo-, regio-, and
Scheme 3. Diversity Enabled by Boron Substituent
entry
acetate
L
yielda (%)
eeb (%)
1
2
3
4
5
6
7
8
9
(rac)-1a
(rac)-1a
(rac)-1a
(rac)-1a
(rac)-1a
(S)-1a
PPh3
77c
80
84
85
25
76
80
80
80
0
(S,S)-L1
(R,R)-L2
(R,R)-L3
(R,R)-L4
PPh3
78 (S)
54 (R)
30 (S)
11 (R)
88 (S)
97 (S)
99 (S)
99 (S)
stereoselective manner. The resulting products could be
used in a large variety of transformations using further
reactions of the pinacol boronate moiety. These exciting
results have compelled us to investigate the scope and
detailed influence of boron in this reaction, and results will
be reported in due course.
(S)-1a
(R,R)-L1
(S,S)-L1
(Rac)-L1
(S)-1a
(S)-1a
a Isolated yield after purification by flash chromatography. b deter-
mined by HPLC using Daicel AS-H column; major enantiomer is noted
between parentheses. c Pd(OAc)2 was used instead of [Pd(allyl)Cl]2.
Note Added after ASAP Publication. A text correction
was made in the third paragraph in the version reposted
July 12, 2011.
meaning that the stereochemical outcome of the reaction is
mostly driven by the substrate (Table 3, entry 7). This was
confirmed when a racemic mixture of the DACH-phenyl
Trost ligand was used, leading to a 98.6% ee indicating a
complete stereoretention at the acetate center (Table 3,
Supporting Information Available. Experimental pro-
1
cedures, characterization data, H, 13C, and 11B NMR
spectra, and HPLC traces. This material is available free
of charge via the Internet at http://pubs.acs.org.
(17) Tao, C.-Z.; Cui, X.; Li, J.; Liu, A.-X.; Liu, L.; Guo, Q.-X.
Tetrahedron Lett. 2007, 48, 3525–3529.
(16) Loiseleur, O.; Elliott, M. C.; von Matt, P.; Pfaltz, A. Helv. Chim.
Acta 2000, 83, 2287.
Org. Lett., Vol. 13, No. 15, 2011
4135