Cyanative alkene-aldehyde coupling: Ni(0)-NHC-Et2AlCN mediated chromanol synthesis with high cis-selectivity at room temperature

Article abstract of DOI:10.1039/b918626c

Described are several classes of Ni(0) mediated cyanative alkene-aldehyde coupling reactions, providing 6-membered cores, which complement existing cyclization technology in several respects. Et₂AlCN was used as both a cyclization accelerator and CN source. The NHC ligand may have a positive effect in differentiating reductive elimination and syn-β-hydride elimination.

Full text of DOI:10.1039/b918626c

View Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Cyanative alkene–aldehyde coupling: Ni(0)–NHC–Et2AlCN mediated
chromanol synthesis with high cis-selectivity at room temperaturew
Chun-Yu Ho
Received (in Cambridge, UK) 9th September 2009, Accepted 26th October 2009
First published as an Advance Article on the web 13th November 2009
DOI: 10.1039/b918626c
Described are several classes of Ni(0) mediated cyanative alkene–
aldehyde coupling reactions, providing 6-membered cores, which
complement existing cyclization technology in several respects.
Et2AlCN was used as both a cyclization accelerator and CN source.
The NHC ligand may have a positive effect in differentiating
reductive elimination and syn-b-hydride elimination.
via oxanickelacycle formation, providing synthetically useful
unsaturated cyclopentanols (eqn (3), Scheme 1), allylic or
homoallylic alcohols through a b-hydride elimination step.7,8
These new findings triggered exploration for a cis-cyclization
strategy that would give rise to cyclohexanol-type structures.
Herein I describe a novel synthetic strategy for their
construction with high diastereoselectivity using Ni(0)–NHC9 and
Nagata’s reagent (Et2AlCN) (eqn (4)).10a-c This strategy not only
provides the first coupling reaction allowing the construction of
6-membered ring alcohols from monoenals with high cis-diastereo-
selectivity, but also the first cyanative alkene–aldehyde coupling,
with concomitant formation of 2 new C–C bonds on those alkenes
regioselectively. Chromanol-type structures (an oxygen-substituted
6-membered cycle fused with an aromatic ring) are the core of
many biologically important molecules;12 2-allyloxybenzaldehyde
S1a was therefore selected as a test substrate for the cyanative
alkene–aldehyde coupling, intended to provide a facile access to
those compounds. Also, with the incorporation of highly versatile
nitrile functionality, it should offer a valuable common inter-
mediate for the synthesis of derivatives when necessary (Scheme 1).
Systematic investigation led to identification of the optimal
ligand, solvent and source of nitrile. Initial screening of cyanating
reagents was conducted by using Ni(cod)2, 1,3-bis(2,6-diisopropyl-
phenyl)imidazol-2-ylidene (IPr) and toluene. Silyl cyanides
(TMS or TBSCN) were found to be not effective in promoting
the cyclization reaction, and large amounts of a-cyanohydrin and
deallylation products were observed. Inspired by the pioneering
work reported by Ogoshi et al.7b using AlMe3 in the cyclization
of monoenones providing allylic alcohols (eqn (3)) and the
fundamental studies on Lewis acidity/nucleophilicity of organo-
aluminiums reported by Lehmkuhl and Snider et al.,13 Et2AlCN
was tested as both a cyanating reagent and a promoter for the
oxidative cyclization. This new combination allowed the highly
cis-diastereoselective cyclization of a monoenal for the first time,
providing chromanol P1a in high yield simply at room tempera-
ture. In situ generation of Me2AlCN for the coupling reaction
from TMSCN and AlMe3 according to a literature procedure was
found to be less effective.10d Replacing the NHC ligand with bulky
Cy3P produced an observable amount of desired coupling
product, but in less than 10% yield (cf. eqn (3)). Similarly, other
phosphorus ligands employed resulted in a-cyanohydrin forma-
tion and serious deallylation. The use of ethereal solvents such as
THF and diethyl ether was also unsuccessful.11 After further
optimizing the physical parameters including reaction
temperature and rate of addition, a series of substituted
chromanol precursors were subjected to the optimized
conditions, and the results are summarized in Table 1.
Reductive radical cyclizations of alkenes with carbonyl groups are
one of the most frequently employed cyclization strategies to
construct cyclic alcohols; they generally favour the thermo-
dynamically more stable trans-isomers.1 For the selective
construction of the less stable cis-isomers, titanocene complexes
have been employed, via a mechanistically different cyclization
strategy involving an oxatitanacycle formation step.2,3 Various
cyclopentanols can be constructed in this way with excellent
cis-diastereoselectivity (eqn (1), Scheme 1), however, precursors
that would give rise to cyclohexanol-type structures (both all
carbon or heterosubstituted) do not cyclize at all using these
complexes.4 Buchwald et al. first demonstrated that the
6-membered cis-cyclization requires an alternative means to
achieve. Carbonylative cyclization mediated by Cp2Ti(CO)2 allows
the 6-membered cyclization of an monoenone and provides the
corresponding cis-g-butyrolactone under thermally induced
reductive elimination conditions (eqn (2), Scheme 1, 100 1C, 40%
based on enone, or 80% based on Ti).5 However, the cis-cyclization
strategy that can join the corresponding monoenal together with
high selectivity remains elusive.6 Recent discoveries have
shown that Ni(0)–phosphine complexes can serve in a similar
capacity in terms of joining monoene and carbonyl together
Scheme 1 Cyclization strategies for monoenals and enones.
Center of Novel Functional Molecules, Department of Chemistry,
The Chinese University of Hong Kong, ShatinNT Hong Kong SAR.
E-mail: jasonhcy@cuhk.edu.hk; Fax: +(852) 2603-5057
w Electronic supplementary information (ESI) available: Experimental
details and data for new compounds. See DOI: 10.1039/b918626c
Although the excellent cis-selectivity was found to be general,
the results showed that substituents on the aromatic ring had a
ꢀc
This journal is The Royal Society of Chemistry 2010
466 | Chem. Commun., 2010, 46, 466–468

View Online
Table 1 Diastereoselective nitrile–alkene–carbonyl couplinga
Entry
1
Substrate
Product
X =
Y =
Yield (%)b
a
b
c
d
e
f
g
h
i
O
H
5-Cl
5-NO2
5-Me
5-CH2NMePh
5-CH2OMe
5-CH2OiPr
5-CH2Nopol
5-OMe
4-OMe
3-OMe
H
92 (87)c
20
0
87
84
88
90
88
O
O
O
O
O
O
O
O
O
O
NTs
CH2
P1
S1
83
70 (83)d
60 (77)d
56 (71)d
34 (68)e
j
k
H
2
a
Me
59 (77 : 23)f
62 (90 : 10)f,g
55 (78 : 22)f
55 (90 : 10)f,g
S2
P2
b
n-Pent
a
See (eqn (4)) and ESI for details. Standard conditions: Ni(cod)2 (0.1 mmol), IPr (0.11 mmol), toluene (2.0 mL), cyclization precursor (0.1 mmol).
b
At 0 1C, 1 M Et2AlCN in toluene (0.2 mmol) was added dropwise. The mixture was stirred for 2 h at 0 1C and 18 h at r.t. Isolated yield and
average of at least 2 runs. 3 times the scale, and using 3.0 mL of toluene. Isolated yield obtained using 1.2 equivalents of Ni(cod)2 and IPr.
c
d
e
f
g
Yield based on starting material recovery. Combined yield of two isomers (cis-trans : cis-cis). Reaction run at À5 1C in the first 2 h.
remarkable effect on the coupling yield (entries 1a–d). Several
monoenals bearing different functional groups were tested for their
compatibility. Luckily, alkyl/benzyl/aryl amines (entry 1e) and
ethers (entries 1f–i) are tolerated, and no nitrile exchanged product
was observed. A 6-membered nitrogen heterocycle can also be syn-
thesized in an analogous way (a hydroquinolinol, entry 1j). While
Ti complexes were reported to be unable to cyclize 6-membered all-
carbon precursors even at elevated temperature, a hydronaphthol
can be synthesized in an observable amount by using the Ni–IPr
complex at room temperature (entry 1k). It is noteworthy that a
monoenal having a simple alkyl substituent on the alkene a-position
is tolerated by this method (entry 2), allowing 3 consecutive
stereogenic centers to be set up on the chromanol, e.g. calanolide
core,11 in a single operation with good diastereoselectivity. Addi-
tional conformational strains experienced during the cyclization
step may account for the lower coupling yield. However, Ni–IPr
was found to be less effective in cyclizing monoenones/5-membered
precursors when compared to Ti methods (see ESIw).
by Ogoshi,7b consistent with the role of ZnMe2 proposed by
Montgomery and Schlegel in other cyclization processes,14 and
the coupling results here. The Et2AlCN acts as a Lewis acid
similar to the Me3Al to facilitate the oxanickelacycle formation.
Et2AlCN here also acted as a nitrile source and the more diffused
sp hybridized CN group is then transferred to the Ni-center
preferentially over the two sp3 hybridized Et groups. The
relatively high stability of the Ni–NHC bond and the sterically
demanding nature of the NHC ligand remarkably slowed down
formal syn-b-hydride elimination (eqn (3)),7a,b,8 and finally allows
the reductive elimination of the cyanation product.
The cis-trans-selectivity observed in the case of the substrate
having an a-branched olefin can be attributed to the preferential
formation of an energetically more favorable oxanickelacycle,
avoiding the steric interaction between the olefin substituent and
the Ni–NHC complex. No allylic alcohol was observed in all
cases examined and the treatment of the independently synthe-
sized allylic alcohol with the standard conditions did not give any
desired product, also supporting the proposed mechanism.
Although Et2AlCN is Lewis acidic in nature, another possible
pathway that may involve the attack of another equivalent of
Et2AlCN on the sterically hindered NiII–CH2 cannot excluded.10e
Interrupting the nitrile transfer with MeOH or O2 at À15 1C
allowed the related hydrogenated or hydroxylated products to be
isolated in small amounts, providing indirect evidence for the
reaction sequence (Scheme 3).15
A possible pathway for product formation with high
cis-selectivity is depicted in Scheme 2. This is largely based on
the reported crystal structure of Ni(PCy3)–Me3Al with d,o-enone
Several notable observations about this new coupling strategy
deserve further comment. (1) Formation of a p-allyl–metal complex
Scheme 2 Proposed mechanistic framework and NHC effect on
reductive elimination versus the b-hydride elimination pathway.
Scheme 3 Experiments to support the proposed reaction sequence.
Chem. Commun., 2010, 46, 466–468 | 467
ꢀc
This journal is The Royal Society of Chemistry 2010

View Online
2 Carbocycle synthesis via titanacycles: (a) S. L. Buchwald and R. B.
Nielsen, Chem. Rev., 1988, 88, 1047; (b) E. Negishi and T. Takahashi,
Acc. Chem. Res., 1994, 27, 124; review on heterocyclic synthesis:
(c) I. Nakamura and Y. Yamamoto, Chem. Rev., 2004, 104, 2127.
3 Advances: (a) N. M. Kablaoui and S. L. Buchwald, J. Am. Chem.
Soc., 1995, 117, 6785; (b) W. E. Crowe and M. J. Rachita, J. Am.
Chem. Soc., 1995, 117, 6787; (c) S. K. Mandal, S. R. Amin and
W. E. Crowe, J. Am. Chem. Soc., 2001, 123, 6457.
Scheme 4 Change in Et2AlCN chemical activity due to the Ni–IPr
complex.
4 The groups of Whitby, Crowe and Buchwald also found that
substrates that would give rise to cyclohexanols do not cyclize at
all: (a) D. F. Hewlett and R. J. Whitby, J. Chem. Soc., Chem.
Commun., 1990, 1684; (b) ref. 3b; (c) N. M. Kablaoui and
S. L. Buchwald, J. Am. Chem. Soc., 1996, 118, 3182.
5 For a breakthrough in the 6-membered cis-cyclization of an enone
(2-allyloxy acetophenone as the only example, no enal example),
see: N. M. Kablaoui, F. A. Hicks and S. L. Buchwald, J. Am.
Chem. Soc., 1997, 119, 4424.
Fig. 1 Complementarity of chromanol synthesis from monoenals/
6 The Ti(OiPr)3Cl/c-C5H9MgCl system mediated imide–monoene
coupling to give 6-membered acylaminals (53%, relative stereo-
chemistry not specified, presumably cis by comparison): J. Lee,
J. D. Ha and J. K. Cha, J. Am. Chem. Soc., 1997, 119, 8127.
7 Intramolecular examples: (a) S. Ogoshi, M. Oka and H. Kurosawa,
J. Am. Chem. Soc., 2004, 126, 11802; (b) S. Ogoshi, M. Ueta,
T. Arai and H. Kurosawa, J. Am. Chem. Soc., 2005, 127, 12810;
(c) M. Murakami and S. Ashida, Chem. Commun., 2006, 4599.
8 Intermolecular examples: (a) C.-Y. Ho, H. Ohmiya and
T. F. Jamison, Angew. Chem., Int. Ed., 2008, 47, 1893;
(b) C.-Y. Ho and T. F. Jamison, Angew. Chem., Int. Ed., 2007,
46, 782; (c) S.-S. Ng, C.-Y. Ho and T. F. Jamison, J. Am. Chem.
Soc., 2006, 128, 11513; (d) C.-Y. Ho, S.-S. Ng and T. F. Jamison,
J. Am. Chem. Soc., 2006, 128, 5362; (e) S.-S. Ng and T. F. Jamison,
J. Am. Chem. Soc., 2005, 127, 14194.
enones, and the derivatives accessible by this work.
upon the addition of the electron-rich early transition metal center
to allyl ethers was often observed, and they are considered to be
incompatible functional groups. Interestingly, allyl ethers and
amines were tolerated here, probably due to the coordination of
the adjacent aldehyde functionality with the metal center and the
mild conditions employed. (2) The addition of M–CN to polarized
carbon–heteroatom multiple bonds is known to proceed readily
even at very low temperature (e.g. a-cyanohydrin formation), and
the results here showed that these pathways can be modulated by
the action of a Ni(0)–IPr complex, providing the first example of
the nitrile group in R2AlCN transferring preferentially to an
electronically neutral CQC in the presence of a polarized CQO
functionality (Scheme 4). A syn-b-hydride elimination process was
reported to occur readily after the oxanickelacycle opening.7,8
Interestingly, the use of NHC may favour a reductive elimination
(and hydrogenation/oxidation) over the b-hydride elimination
pathway, preserving the two consecutive stereocenters for the first
time. This effect may have broader implications for the cyclization
using Ni–NHC complexes. (4) Whereas typical radical cyclization
conditions favour the trans-isomer and are terminated by H, the
high cis-selectivity and the incorporation of highly versatile CN
here are noteworthy.16–18 A collection of structurally related
chromanol derivatives should be able to be reached by simple
manipulation (e.g. reduction/hydrolysis, see the ESIw) and may
have an advantage in biological screening (Fig. 1).
9 Recent reviews of NHCs in transition-metal catalysis:
(a) N. Marion and S. P. Nolan, Acc. Chem. Res., 2008, 41, 1440;
(b) S. Wurtz and F. Glorius, Acc. Chem. Res., 2008, 41, 1523.
¨
10 (a) W. Nagata and M. Yoshioka, Tetrahedron Lett., 1966, 7, 1913;
(b) W. Nagata and M. Yoshioka, Org. React., 1977, 25, 255;
(c) M. North, D. L. Usanov and C. Young, Chem. Rev., 2008,
108, 5146; in situ generationof Me2AlCN: (d) A. Westwood and
D. Nicholls, Inorg. Chim. Acta, 1996, 245, 97; no cyanoalumina-
tion example for monoenes using R2AlCN was found. The
addition of organyl groups on an Al center to monoenes generally
occurs at the 2-position: (e) M. Oishi, Sci. Synth., 2004, 7, 261.
11 Efforts were made to render the coupling catalytic, however, unlike
ref. 8, using amines or phosphites or both as additives was found
not to be successful. All the reagents are commercially available,
which may ease the complications in preparation.
12 Examples included anti-HIV-1 calanolide and the inophyllum family.
A recent discovery showed that structural modification of chromanol
may provide better inhibitory activity, for leading references see:
(a) P. P. Deshpande, F. Tagliferri, S. F. Victory and D. C. Baker,
J. Org. Chem., 1995, 60, 2964; (b) B. M. Trost and F. D. Toste, J. Am.
Chem. Soc., 1998, 120, 9074; (c) T. Ma, L. Liu, H. Xue, L. Li, C. Han,
L. Wang, Z. Chen and G. Liu, J. Med. Chem., 2008, 51, 1432.
13 (a) H. Lehmkuhl, Angew. Chem., 1963, 75, 1090; (b) B. B. Snider,
D. J. Rodini, M. Karras, T. C. Kirk, E. A. Deutsch, R. Cordova
and R. T. Price, Tetrahedron, 1981, 37, 3927; for a review see:
(c) S. Saito, Sci. Synth., 2004, 7, 95.
In summary, cyanative alkene–aldehyde coupling has been
achieved by a novel use of Ni–NHC/Et2AlCN, applied to
6-membered heterocycle synthesis, providing a cyclization
strategy complementary to existing technology and new
inspirations in both Ni and Al chemistry. With this
fundamental molecular transformation demonstrated, further
development of related systems is now underway.
14 (a) J. Montgomery, Angew. Chem., Int. Ed., 2004, 43, 3890;
(b) R. M. Moslin, K. M. Moslin and T. F. Jamison, Chem.
Commun., 2007, 4441; (c) Modern Organonickel Chemistry;
ed. Y. Tamaru, Wiley-VCH, Weinheim, Germany, 2005.
15 E. Nakamura, H. Oshino and I. Kuwajima, J. Am. Chem. Soc.,
1986, 108, 3745; Y. Sato, T. Takanashi and M. Mori, Organo-
metallics, 1999, 18, 4891.
Support for this work was provided by the Center of Novel
Functional Molecule, The Chinese University of Hong Kong,
Incentive Research Scheme (4440138) and the Hong Kong
General Research Fund (2160344).
16 Chromanol synthesis using Bu3SnH conditions: J. Bentley, P. A. Nilsson
and A. F. Parsons, J. Chem. Soc., Perkin Trans. 1, 2002, 1461 (from S1a,
24%, trans :cis = 5 : 1, along with 62% reduction).
Notes and references
1 Examples using lanthanide: (a) G. A. Molander, in The Chemistry of
the Metal–Carbon Bond, ed. F. R. Hartley, J. Wiley & Sons, New York,
1989, vol. 5, ch. 8, p. 319; O-stannyl ketyl radical: (b) A. L. J. Beckwith,
D. M. O’Shea and D. H. Roberts, J. Chem. Soc., Chem. Commun.,
1983, 1445; (c) E. J. Enholm and K. S. Kinter, J. Am. Chem. Soc.,
1991, 113, 7784; electroreductive: (d) J. E. Swartz, T. J. Mahachi and
E. Kariv-Miller, J. Am. Chem. Soc., 1988, 110, 3622.
17 Using titanocene radicals, in situ generated from Cp2TiCl2/Zn
dust: S. Jana and S. C. Roy, Tetrahedron Lett., 2006, 47, 5949
(from S1a, 72%, stereoselectivity not determined).
18 Reductive vinylnitrile–aldehyde cyclization also gives the trans-product:
electroreductive strategy: (a) J. A. Miranda, C. J. Wade and R. D.
Little, J. Org. Chem., 2005, 70, 8017; SmI2-mediated: (b) M. Tamiya,
C. Jager, K. Ohmori and K. Suzuki, Synlett, 2007, 5, 780.
ꢀc
This journal is The Royal Society of Chemistry 2010
468 | Chem. Commun., 2010, 46, 466–468