Mechanistic insight into copper-catalysed allylic substitutions with bis(triorganosilyl) zincs. Enantiospecific preparation of α-chiral silanes

Article abstract of DOI:10.1039/b606589a

Isotopic desymmetrisation, as well as (stereo)chemical correlation, has illuminated significant aspects of the σ-π-σ mechanism of copper-catalysed allylic substitution reactions: an enantiospecific and regioselective access to α-chiral silanes is presente

Full text of DOI:10.1039/b606589a

View Article Online / Journal Homepage / Table of Contents for this issue
COMMUNICATION
www.rsc.org/chemcomm | ChemComm
Mechanistic insight into copper-catalysed allylic substitutions with
bis(triorganosilyl) zincs. Enantiospecific preparation of a-chiral
silanes{
Eric S. Schmidtmann and Martin Oestreich*
Received (in Cambridge, UK) 9th May 2006, Accepted 4th July 2006
First published as an Advance Article on the web 28th July 2006
DOI: 10.1039/b606589a
Isotopic desymmetrisation, as well as (stereo)chemical correla-
tion, has illuminated significant aspects of the s–p–s mechanism
of copper-catalysed allylic substitution reactions: an enantio-
specific and regioselective access to a-chiral silanes is presented.
Allylic silanes offer a versatile synthetic chemistry.1 Among the
established methods for their direct preparation, the substitution of
allylic substrates with (Me2PhSi)2CuLi?LiCN is particularly
Scheme 1 Allylic substitution at a ‘‘symmetric’’ cyclic substrate.
effective, if the regio- and diastereoselectivity are controllable
elements.2 Extensive investigations, mainly by Fleming et al.,3 but
subsequent regular acylation. We intentionally chose these
also by Kitching et al.4 and Clive et al.,5 revealed the subtle
substrates as anti-substitution is well documented8 and therefore
influence of solvent mixtures and the Me2PhSiLi–CuX ratio (X =
the lack of regiocontrol would directly correlate with racemisation.
CN or I) on the isomeric distribution. Lipshutz et al. later showed
Indeed, standard reaction conditions{ afforded allylic silane (R)-2
that the allylic substitution of vinylic epoxides is feasible using
as an almost racemic mixture (1A2, Scheme 1).
zincate (Me2PhSi)Me2ZnLi in the presence of catalytic amounts of
2
In order to further substantiate this finding, we prepared H-
cuprate Me2CuLi?LiCN.6 As we reported recently, the reagent–
labelled syn-3a and syn-3b (regiochemical probe) substituted with a
catalyst combination (Me2PhSi)2Zn and CuI (5.0 mol%)7 enables
methyl group (stereochemical probe) (Scheme 2). As expected,
the facile copper-catalysed allylic substitution of selected allylic
substitution of both benzoate syn-3a and carbamate syn-3b
esters and carbamates with perfect regio- and/or diastereoselec-
produced a mixture of regioisomers anti-4 and anti-5 with an
tivity.8 Conversely, the enantiospecific preparation by these
unchanged (inverted) diastereomeric ratio8 (Table 1, entries 1 and
procedures of synthetically valuable a-chiral allylic silanes,9 that
2). Accordingly, the copper-mediated allylic substitution reaction
is allylic silanes with silicon attached to a stereogenic carbon, is
using Me2PhSiCu?LiCN (Table 1, entries 3 and 4), the supposed
catalytically-active species generated from (Me2PhSi)2Zn and CuI
(5.0 mol%),7b and (Me2PhSi)2CuLi?LiCN (Table 1, entries 5 and 6)
gave identical results.
considerably more complex.3c,d The mechanism of these copper-
mediated and, in particular, copper-catalysed allylic silylation
reactions is still vague.10
In the light of the ready availability of chiral, non-racemic allylic
These experiments indicate that the copper-catalysed transfer of
alcohols and cognate esters, as well as carbamates, we targeted the
nucleophilic silicon from (Me2PhSi)2Zn onto cyclic allylic sub-
direct access to a-chiral allylic silanes by our copper-catalysed
strates might involve symmetric (achiral) p allyl copper(III)
allylic substitution route with bis(triorganosilyl) zinc,
intermediates (vide infra).
(Me2PhSi)2Zn. In this communication, we present a systematic
We then addressed the allylic substitution of acyclic substrates,
investigation of its stereochemical course by a combination of
isotopic desymmetrisation11 and chemical correlation. A mechan-
in which the intermediacy of symmetric p allyl copper(III)
complexes is also possible. Since 1,3-diphenylallyl esters failed to
undergo substitution,14 we decided to utilise alkyl-substituted
ism, not always likely to proceed through conventional s–p–s
allylic isomerisation,12 evolves from this study. Enantiospecific
benzoate 6a and carbamate 6b (Scheme 3 and Scheme 4). This
allylic displacement of the leaving group by an anti-SN reaction
time, the investigation commenced with the allylic substitution
pathway was accomplished.
Our investigation started with the copper-catalysed allylic
substitution of enantioenriched cyclic (S)-1a and (S)-1b
(Scheme 1), which were each prepared by palladium-catalysed
deracemisation of the corresponding allylic carbonate13 and
Organisch-Chemisches Institut, Westfa¨lische Wilhelms-Universita¨t,
Corrensstraße 40, D-48149 Mu¨nster, Germany.
E-mail: martin.oestreich@uni-muenster.de; Fax: +49 (0)251 83-36501;
Tel: +49 (0)251 83-33271
{ Electronic supplementary information (ESI) available: Characterisation
data for all new compounds. See http://dx.doi.org/10.1039/b606589a/
2
Scheme 2 Mechanistic insight by H-labelling (part 1).
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 3643–3645 | 3643

View Article Online
Table 1 Comparative survey of cuprate reagents
Entry
Compound
R
Cuprate reagent
Yield (%)a
rs (4 : 5) (%)b
1
2
3
4
5
6
a
3a
3b
3a
3b
3a
3b
C(O)Ph
C(O)NHPh
C(O)Ph
C(O)NHPh
C(O)Ph
C(O)NHPh
(Me2PhSi)2Zn–CuI (5.0 mol%)
(Me2PhSi)2Zn–CuI (5.0 mol%)
Me2PhSiCu?LiCN
77
70
62
72
73
78
50 : 50
50 : 50
50 : 50
50 : 50
50 : 50
50 : 50
Me2PhSiCu?LiCN
(Me2PhSi)2CuLi?LiCN
(Me2PhSi)2CuLi?LiCN
b
Yield of analytically pure product after purification by flash chromatography. The ratio of regioisomers was determined from the 1H NMR
spectra by integration of the baseline-separated resonance signals prior to purification.
2
Scheme 3 Mechanistic insight by H-labelling (part 2).
Scheme 5 Allylic substitution at a ‘‘non-symmetric’’ acyclic substrate.
Scheme 4 Allylic substitution at a ‘‘symmetric’’ acyclic substrate.
2
reaction of H-labelled racemic rac-6-d1. In sharp contrast to the
cyclic system (Scheme 2), allylic transfer of the silicon nucleophile
proceeded with moderate regioselectively in the case of the acyclic
system (rac-6-d1Arac-7-d1, Scheme 3).15 This unexpected observa-
tion was corroborated in the analogous substitution reactions of
enantioenriched substrates (R)-6, which were again accessed by
palladium-catalysed deracemisation.13 Importantly, enantioen-
riched a-chiral allylic silane (S)-7 was obtained with only a minor
loss of stereochemical information (6A7, Scheme 4).
This remarkable conservation of the absolute configuration
(anti-SN mechanism), occasionally termed a memory effect, was
also found in rhodium-16 and palladium-catalysed17 allylic
alkylation reactions, yet it was interpreted by entirely different
explanations. Similar to the rhodium catalysis reported by Evans
et al.,16 the copper-catalysed allylic substitution of acyclic
substrates might favour non-symmetric (chiral) s allyl or [s + p]
enyl18 rather than p allyl copper(III) intermediates (vide infra).
In the enantiospecific substitution of allylic substrates with
different substituents, the issues of regio- and enantioselectivity are
separated. For this, we prepared (S)-8a, (S)-8b and (R)-8c by
enzymatic kinetic resolution of the parent alcohol.19 To our
delight, allylic substitution provided a-chiral allylic silanes (R)-9
and (S)-9 with excellent regio- and enantioselectivity, independent
of the leaving group (8A9, Scheme 5).20 This inversion (formal
Scheme 6 Unified mechanistic rationale (ox. add. = oxidative addition,
red. elim. = reductive elimination, L = solvent or counter-ion).
anti-SN mechanism) at the silicon-bearing stereogenic carbon was
unambiguously secured by chemical correlation in two synthetic
operations:21 racemisation-free hydrogenation of the double bond
From these experimental findings, we deduce the unified mecha-
in (S)-9 followed by oxidative degradation of the carbon–silicon
bond under retention of configuration (Fleming oxidation22).
nistic picture outlined in Scheme 6. We propose that allylic sub-
strate A oxidatively adds to catalytically-active monosilylcopper
3644 | Chem. Commun., 2006, 3643–3645
This journal is ß The Royal Society of Chemistry 2006

View Article Online
Me2PhSiCu?ZnX2 (X = Me2PhSi, Cl and I),7b thereby forming the
[s + p] enyl copper(III) transient intermediate B;18 the related s
allyl copper(III) complex is not shown. This oxidative addition
proceeds in an anti-SN fashion, that is, inversion and no a/c
transposition (AAB). Then, reductive elimination (BAD) and
conventional s–p–s isomerisation12 (BAC/C9AB9) are possible
reaction pathways; again, the s allyl copper(III) complex,
corresponding to B9, is not shown. Depending on the individual
structural features and the substitution pattern of the allyl
fragment, two scenarios might explain our observations
(Scheme 6):23
(Na2SO4), the solvents were evaporated under reduced pressure and the
crude product purified by flash chromatography on silica gel using
cyclohexane as the solvent. All new compounds gave satisfactory
characterisation data (ESI{).
1 L. Chabaud, P. James and Y. Landais, Eur. J. Org. Chem., 2004,
3173–3199.
2 R. K. Dieter, in Modern Organocopper Chemistry, ed. N. Krause, Wiley-
VCH, Weinheim, 2002, pp. 79–144.
3 (a) I. Fleming and T. W. Newton, J. Chem. Soc., Perkin Trans. 1, 1984,
1805–1808; (b) I. Fleming and A. P. Thomas, J. Chem. Soc., Chem.
Commun., 1985, 411–413; (c) I. Fleming and A. P. Thomas, J. Chem.
Soc., Chem. Commun., 1986, 1456–1457; (d) I. Fleming, D. Higgins,
N. J. Lawrence and A. P. Thomas, J. Chem. Soc., Perkin Trans. 1, 1992,
3331–3349; (e) I. Fleming and N. K. Terrett, J. Chem. Soc., Perkin
Trans. 1, 1998, 2645–2649; (f) I. Fleming and D. Higgins, J. Chem. Soc.,
Perkin Trans. 1, 1998, 2673–2678.
4 (a) B. Laycock, W. Kitching and G. Wickham, Tetrahedron Lett., 1983,
24, 5785–5788; (b) B. Laycock, I. Maynard, G. Wickham and
W. Kitching, Aust. J. Chem., 1988, 41, 693–700.
5 D. L. J. Clive, C. Zhang, Y. Zhou and Y. Tao, J. Organomet. Chem.,
1995, 489, C35–C37.
N Path 1 (anti-SN/anti-SN9 and a-adduct/c-adduct): If [s + p] enyl
complex B is cyclic, and R1 and R2 are not capable of establishing
p-conjugation, isomerisation (BAC/C9) will be faster relative to the
rate of reductive elimination (BAD). Achiral p allyl complexes C
and C9 will be in equilibrium with B and B9, which will eventually
suffer reductive elimination (BAD and B9AD9). With R1 = R2 =
–(CH2)3–, racemisation is observed (1A2, Scheme 1).
N Path 2a (anti-SN and a-adduct): If [s + p] enyl complex B is
acyclic, and R1 and R2 are not capable of establishing
p-conjugation, reductive elimination (BAD) will be slightly faster
relative to the rate of isomerisation (BAC/C9). With R1 = R2 =
Me, partial erosion of the stereochemical information, and hence
overall inversion of configuration, is observed (6A7, Scheme 4).
Identical behaviour is seen for a related acyclic substrate (R1 =
R2 = Et).
6 B. H. Lipshutz, J. A. Sclafani and T. Takanami, J. Am. Chem. Soc.,
1998, 120, 4021–4022.
7 (a) M. Oestreich and B. Weiner, Synlett, 2004, 2139–2142; (b) G. Auer
and M. Oestreich, Chem. Commun., 2006, 311–313; (c) G. Auer,
B. Weiner and M. Oestreich, Synthesis, 2006, 2113–2116.
8 M. Oestreich and G. Auer, Adv. Synth. Catal., 2005, 347, 637–640.
9 Representative stereoselective syntheses of a-chiral allylic silanes: (a)
Catalyst control: T. Hayashi, M. Konishi, Y. Okamoto, K. Kabeta and
M. Kumada, J. Org. Chem., 1986, 51, 3772–3781; (b) Reagent control:
W. R. Roush and P. T. Grover, Tetrahedron, 1992, 48, 1981–1998; (c)
Substrate control: J. S. Panek and T. D. Clark, J. Org. Chem., 1992, 57,
4323–4326; (d) Substrate control: M. Suginome, A. Matsumoto and
Y. Ito, J. Am. Chem. Soc., 1996, 118, 3061–3062; (e) Substrate control:
J. H. Smitrovich and K. A. Woerpel, J. Org. Chem., 2000, 65,
1601–1614.
10 S. Mori and E. Nakamura, in Modern Organocopper Chemistry, ed.
N. Krause, Wiley-VCH, Weinheim, 2002, pp. 315–346.
11 G. C. Lloyd-Jones, Synlett, 2001, 161–183.
12 B. M. Trost and D. L. Van Vranken, Chem. Rev., 1996, 96, 395–422.
13 B. J. Lu¨ssem and H.-J. Gais, J. Am. Chem. Soc., 2003, 125, 6066–6067.
14 Instead, (E)-1,3-diphenylpropene was isolated in trace amounts (y10%
yield), indicating the formation of a stable exo,exo p allyl copper(III)
intermediate.
15 E,Z ratios were determined from 1H NMR spectra before and after
purification by flash chromatography on silica gel. As verified by
comparison with reported data,9e only the E and not the Z isomer was
detectable.
16 P. A. Evans and J. D. Nelson, J. Am. Chem. Soc., 1998, 120, 5581–5582.
17 (a) B. M. Trost and R. C. Bunt, J. Am. Chem. Soc., 1996, 118, 235–236;
(b) T. Hayashi, M. Kawatsura and Y. Uozumi, J. Am. Chem. Soc.,
1998, 120, 1681–1687; (c) G. C. Lloyd-Jones and S. C. Stephen, Chem.–
Eur. J., 1998, 4, 2539–2549; (d) A. J. Blacker, M. L. Clarke, M. S. Loft
and J. M. J. Williams, Org. Lett., 1999, 1, 1969–1971; (e) L. Acemoglu
and J. M. J. Williams, Adv. Synth. Catal., 2001, 343, 75–77.
18 M. Yamanaka, S. Kato and E. Nakamura, J. Am. Chem. Soc., 2004,
126, 6287–6293.
19 S. Maier and U. Kazmaier, Eur. J. Org. Chem., 2000, 1241–1251.
20 (a) An isomeric mixture of 1-phenyl-2-pentene was detected as a major
by-product, indicating the formation of exo,exo and exo,endo p allyl
copper(III) intermediates; (b) It should be noted that treatment of (R)-8c
(92% ee) with (Me2PhSi)2CuLi?LiCN also afforded (S)-9 (92% ee) with
excellent regiocontrol3d.
N Path 2b (anti-SN and a-adduct): If [s + p] enyl complex B is
acyclic, and either R1 or R2 are capable of establishing
p-conjugation, direct reductive elimination (BAD) is again likely
to be faster than isomerisation (BAC/C9). Even if s–p–s
isomerisation occurs, intermediate p allyl complex C is likely to
predominantly isomerise to the more stable [s + p] enyl complex
B, favouring conjugation. This assumption is supported by a
recent theoretical study by Nakamura et al.18 Reductive elimina-
tion will then occur with retention of configuration (BAD). With
R1 ? R2, R1 = Ph and R2 = Et, overall inversion of configuration
is observed (8A9, Scheme 5).
In conclusion, we have developed an enantiospecific synthesis of
a-chiral allylic silanes by copper-catalysed allylic silylation of
readily available enantioenriched allylic substrates. Isotopic
labelling and assignment of absolute configurations by chemical
correlation have disclosed useful insights, which have clarified the
mechanistic trends for several substrates.23
M. O. is indebted to the Deutsche Forschungsgemeinschaft for
an Emmy Noether Fellowship (2001–2006) and to the Aventis
Foundation for a Karl Winnacker Fellowship (2006–2008). We
thank the Fonds der Chemischen Industrie for additional financial
support and Gerd Fehrenbach for performing the HPLC analyses.
Notes and references
{ General experimental procedure: A suspension of CuI (5.0 mol%) and
THF (1.0 mL) was pre-cooled to 278 uC and treated with bis(dimethyl-
phenylsilyl) zinc8 (1.0 equiv.) via a syringe. The auburn-coloured reaction
mixture was allowed to warm to 0 uC and maintained at this temperature
for 0.5 h. Addition of the allylic substrate (1.0 equiv.) in THF (1.0 mL) was
followed by stirring for 1 h at 0 uC. Upon completion of the reaction, the
reaction mixture was poured into saturated aqueous NH4Cl (5.0 mL) and
the flask rinsed with tert-butyl methyl ether (10 mL). The aqueous phase
was separated and extracted with tert-butyl methyl ether (3 6 10 mL). The
combined organic phases were washed with brine (10 mL). After drying
21 A detailed scheme is provided in the ESI{.
22 I. Fleming, R. Henning, D. C. Parker, H. E. Plaut and P. E. J.
Sanderson, J. Chem. Soc., Perkin Trans. 1, 1995, 317–337.
23 At present, we cannot entirely rule out direct nucleophilic substitution
pathways, as well as a sequence consisting of anti-SN9 (oxidative
addtion) and syn-SN9 (reductive elimination). However, the latter, being
complementary to Paths 2a and 2b, seems unlikely because of the
involvement of a syn-SN9 reductive elimination from an intermediate
[s + p] enyl species18.
This journal is ß The Royal Society of Chemistry 2006
Chem. Commun., 2006, 3643–3645 | 3645