Taming the carboxyl group for directed carbometalation: Observations on the use of anions, dianions and ester enolates

Article abstract of DOI:10.1002/chem.201403294

Carboxylate anions, dianions and ester enolates provide simultaneous protection and activation for directed carbometalation reactions. Advantage can be taken of the bis-carbanionic character of the intermediate for further controlled C-C bond forming reac

Full text of DOI:10.1002/chem.201403294

DOI: 10.1002/chem.201403294
Communication
&
Synthetic Methods
Taming the Carboxyl Group for Directed Carbometalation:
Observations on the Use of Anions, Dianions and Ester Enolates
Sandy Desrat, Philip J. Gray, Matthew R. Penny, and William B. Motherwell*[a]
Dedicated to the pioneering spirit of Professors Jean Normant and Alex Alexakis
Abstract: Carboxylate anions, dianions and ester enolates
provide simultaneous protection and activation for direct-
ed carbometalation reactions. Advantage can be taken of
the bis-carbanionic character of the intermediate for fur-
ther controlled CÀC bond forming reactions.
Carbon–carbon bond forming reactions leading to the con-
struction of a regio- and stereodefined alkene unit[1] are recog-
nised to be an absolutely vital component in the armoury of
the synthetic organic chemist. Within this area, following on
from the pioneering studies of Normant, Alexakis and cowork-
ers,[2] the wide ranging potential of heteroatom directed carbo-
metalation[3] of an alkyne[4] or an allene[5] has burgeoned in
recent years and provides an extremely powerful strategy. The
elegant studies of Marek[6] on directed carbocupration for ex-
ample, provide a perfect illustration that advantage can be
taken, not only of the vinylic carbon–metal bond formed in
the initial carbometalation step, but also of the heteroatom di-
recting group, to forge additional carbon–carbon bonds.
Scheme 1. Enolate directed carbometalation.
The foregoing concept is illustrated in more detail in
Scheme 1 using the simple allenic ester 1a, which was selected
as a suitable substrate for an initial study involving directed
carbocupration via 3 using either of the two oxygen atoms in
the derived ester enolate 2. An additional feature of considera-
ble interest was that the resultant heterobimetallic intermedi-
ate 4 would possess bis-carbanionic character and could there-
fore react, either as an ester enolate and/or as a vinyl cuprate,
in subsequent reactions with selected electrophiles to furnish
polyfunctionalised alkenes 5.
In the first instance, we elected to examine the behaviour of
the lithium ester enolate of ethyl penta-3,4-dienoate[8] 1a in
a stoichiometric carbocupration reaction using the protocol
developed by Normant,[9] which involves mixing copper(I)
iodide with one equivalent of Grignard reagent in diethyl ether
at À308C for 1 h. Examination of the results in Table 1 for
In similar fashion, with the objective of gaining increased ef-
ficiency and versatility in carbon–carbon bond formation, we
were particularly intrigued by the possibility that, through the
simple expedient of enolate anion formation, it would be pos-
sible to use the oxygen atom of a carbonyl group as the di-
recting heteroatom. Such a selection leads to concomitant pro-
tection of the carbonyl group against nucleophilic attack by
main group organometallic reagents together with formation
of one of the preeminent cornerstones for carbon–carbon
bond formation in organic synthesis. To the best of our knowl-
edge, an enolate “protection–activation” strategy has not been
employed in directed carbometalation. We were nevertheless
encouraged by the seminal observation of Normant, Alexakis
and Villieras that carbocupration could be directed by selection
of the (Z)-enol ether geometry of 1-methoxybut-1-ene-3-yne.[7]
Table 1. Carbocupration of 1a using a stoichiometric amount of organo-
copper reagent.
Entry
R
Product
Yield [%][a]
(E)/(Z) ratio[b]
[a] Dr. S. Desrat, Dr. P. J. Gray, Dr. M. R. Penny, Prof. Dr. W. B. Motherwell
Department of Chemistry
1
2
3
4
iPr
Et
nBu
Ph
6
7
8
9
81
93
56
52
80:20
90:10
75:25
30:70
University College London
20 Gordon Street, London WC1H 0AJ (United Kingdom)
Tel : +(44)2076797524
E-mail: w.b.motherwell@ucl.ac.uk
[a] Yield of the isolated products based on starting material 1a. [b] Deter-
mined by 1H NMR spectroscopic analysis of the crude reaction product.
Supporting information for this article is available on the WWW under
http://dx.doi.org/10.1002/chem.201403294.
Chem. Eur. J. 2014, 20, 8918 – 8922
8918
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
a series of organocopper reagents reveals that the
addition step is highly regioselective and that prepa-
ratively useful yields can be obtained, especially
when the aggregated nature of the two reacting
components is considered. It should also be noted
that the stereoselectivity can be strongly influenced
by the nature of the organocopper reagent selected,
as emphasised by the dramatic reversal between the
alkyl (entries 1–3) and aryl substituents (entry 4).
Table 2. Reactivity of the bis-carbanionic intermediate towards electrophiles to form
polyfunctionalised alkenes 7–7c.
Entry E+
Equiv
Product
Yield [%][a] (E)/(Z) ratio[b]
Our attention was then directed towards the excit-
ing possibility of exploiting the bis-carbanionic char-
acter of the carbometalated intermediate 4. In order
to achieve chemoselective reactions, it is of course
mandatory that the rates of reaction of the “ester
enolate component” and the “organocopper compo-
nent” in the complex heterobimetallic intermediate
are sufficiently different for a given electrophile. Ini-
tially, as demonstrated by the isolation of the dideu-
terio compound 7a in high yield through addition of
DCl at the end of the reaction (Table 2, entry 2) it was
possible to confirm the dianionic character of 4. Grat-
ifyingly, selective alkylation at the organocopper site
could also be achieved and the monoalkylated prod-
uct 7b was isolated in moderate yield, even when
using three equivalents of allyl bromide as electro-
phile (Table 2, entry 3). By way of contrast, a prelimi-
1
2
H2O
excess
7
93
90:10
90:10
DCl
excess
7a 96
7b 44
3
4
3 equiv
10:90
90:10
1.1 equiv
7c 21
[a] Yield of the isolated products based on starting material 1a. [b] Determined by
1H NMR spectroscopic analysis of the crude reaction product.
nary experiment involving the addition of one molar equiva-
lent of an aldehyde to the reaction mixture led to selective
trapping of the ester enolate, thus forming the alcohols 7c
(d.r.ꢀ1:1) albeit in low yield (Table 2, entry 4) together with
Table 3. Copper-mediated carbomagnesiation of 1a directed by an eno-
late.
the untrapped carbometallated product
component.
7 as the major
A contemporaneous study of copper-catalysed carbomagne-
siation that involves the intermediacy of a more basic homo-
cuprate species (R2CuM) also revealed several features of inter-
est. Thus, careful optimisation of the reaction conditions
showed that preformation of the ester enolate is essential for
the reaction otherwise addition of the Grignard reagent on the
carbonyl occurred. It also established that premixing 0.1 equiv-
alents of CuI and 1.1 equivalents of Grignard reagent at À208C
in Et2O for 2 h prior to the addition of the ester enolate gave
the best results. Notably, and in contrast to the stoichiometric
carbocupration reaction, the product alkenes 6, 7, 9 and 10,
could be isolated in improved yield irrespective of the
Grignard reagent employed, and with a stereospecific prefer-
ence for formation of the (E) isomer (Table 3).
Entry
R
Product
Yield [%][a]
(E)/(Z) ratio[b]
1
2
3
4
iPr
Et
Ph
Cy
6
7
9
69
71
79
67
100:0
100:0
100:0
100:0
10
[a] Yield of the isolated products based on starting material 1a. [b] Deter-
mined by 1H NMR spectroscopic analysis of the crude reaction product.
Encouraged by this improvement in the stereoselectivity, the
reactivity of the Grignard intermediate was then probed by
treatment of the reaction mixture with DCl in D2O (Scheme 2).
This experiment led to complete deuterium incorporation of
the enolate position a but only 50% of deuterium incorpora-
tion at the vinylic Grignard site g. This result suggested that
only one of the two ester enolate isomers of the highly basic
Grignard intermediate 4, was protonated in situ by the di-iso-
propylamine generated from the use of LDA in the first step.
This hypothesis was simply proven by removal of the di-isopro-
pylamine under high vacuum after reaction with LDA and prior
Scheme 2. Trapping of the reactive species generated by copper mediated
carbomagnesiation of 1a via an enolate.
to the carbometalation step, and led to complete deuterium
incorporation at both sites. In terms of practical utility howev-
er, such additional experimental manipulations were not desir-
Chem. Eur. J. 2014, 20, 8918 – 8922
www.chemeurj.org
8919
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
able, with the deuterated alkene 9a being isolated in only
22% yield.
Table 5. Copper mediated carbomagnesiation of
carboxylate.
1 directed by a -
As a logical consequence of the fact that the ester unit has
two environmentally different oxygen atoms capable of direct-
ing carbometalation, we then explored the behavior of the
symmetrical dilithio dianion of the carboxylic acid 1. As shown
in Table 4, using the conditions optimised for the ester enolate,
the carbometallated product was obtained with good (E)/(Z)
Entry
R
Product
Yield [%][a]
(E)/(Z) ratio[b]
1
2
3
4
5
6
7
Et
nPentyl
iPr
Cy
tBu
Ph
11 a
11 b
11 c
11 d
11 e
11 f
11 g
70
78
95
96
61
74
88
90:10
90:10
95:5
95:5
95:5
Table 4. Copper mediated carbomagnesiation of 1 directed by a dianion.
90:10
95:5
1-Me-1-propenyl
[a] Yield of the isolated products based on starting material 1. [b] Deter-
mined by 1H NMR spectroscopic analysis of the crude reaction product.
Entry
R
Product
Yield [%][a]
(E)/(Z) ratio[b]
1
2
3
4
5
Et
nPentyl
iPr
Cy
tBu
11 a
11 b
11 c
11 d
11 e
38
40
35
34
34
90:10
90:10
95:5
95:5
95:5
[a] Yield of the isolated products based on starting material 1. [b] Deter-
mined by 1H NMR spectroscopic analysis of the crude reaction product.
selectivity. Yields however were moderate, and substrate con-
versions above 50% proved to be problematic, irrespective of
the Grignard reagent employed.
Even although the use of the carboxylate anion might be
considered as the simplest form of protection against nucleo-
philic attack for a carboxylic acid, it has been well known for
over a century that their reaction with main group organome-
tallics is a very well established method for conversion to ke-
tones.[10] We reasoned nevertheless that a copper-catalysed
carbomagnesiation reaction using the allenic carboxylate as
substrate would be the preferred pathway. Examination of the
results for a selection of Grignard reagents illustrates that the
desired alkenes 11 were formed with good (E)/(Z) selectivity
and in very good yield (Table 5). From a practical standpoint,
the reaction is very easy to perform since prior low tempera-
ture formation of the organocopper species is not required,
and, unlike the dianion directed reaction above, complete con-
version of starting material is observed.
Scheme 3. Proposed mechanism of the copper mediated carbomagnesiation
of 1 followed by trapping of the magnesium enolate 14 by DCl.
lation with the Grignard reagent to regenerate the homocup-
rate and form the vinylic Grignard 13. This intermediate is a suf-
ficiently strong base to form the dianion of the carboxylic acid
14, which was then trapped by addition of DCl.
As emphasised by the wide range of electrophiles in Table 6,
advantage can be taken of this dianion intermediate 14 for
further carbon–carbon bond formation in an experimentally
convenient one pot sequence. Thus, a-alkylation with iodo-
methane (entry 1) and formation of b-hydroxy carboxylic acids
from a range of aldehydes and ketones (entries 3–8) all pro-
ceeded in excellent yield, save for the use of paraformaldehyde
(entry 2). The tolerance of halogen functionality in the reac-
tions of a-halo ketones (entries 9–11) is also noteworthy. The
selection of an epoxide as an electrophile leading to a g-lac-
tone (entries 12–14) is however eroded by competing rear-
rangement of the epoxide to a carbonyl compound.
Since further functionalisation of the carbometallated inter-
mediate 4 was a major focus of interest, a copper mediated
carbomagnesiation reaction directed by the carboxylate of
1 was then treated with DCl. To our initial surprise, no deuteri-
um incorporation was observed in the g position, and the a-
deuterated compound 15 was isolated in excellent yield 94%
and with a stereoselectivity of 95:5 in favour of the (E) isomer
(Scheme 3).
The foregoing preliminary results have hopefully illustrated
that the simple concept of anionic protection of the carbonyl
group also leads to concomitant efficiency in directed carbo-
metalation and versatility in controlling further selective
carbon–carbon bond forming reactions. Given the current
wealth of metals that can be used in directed carbometalation,
and the protocols available for regio- and stereospecific eno-
A plausible mechanistic rationale for this regiospecificity is
shown in Scheme 3. Thus, the initial homocuprate generated
in situ reacts with the magnesium carboxylate 1b to give orga-
nocopper intermediate 12, which then undergoes transmetal-
Chem. Eur. J. 2014, 20, 8918 – 8922
www.chemeurj.org
8920
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Table 6. Reactivity of the intermediate generated by copper mediated carbomagnesiation of 1 towards electrophiles to form polyfunctionalised alkenes.
Entry
1
E+
Equiv
Products
Yield [%][a]
82
MeI
3 equiv
16a
16b
2
3
p-formaldehyde
8 equiv
3 equiv
49
16c
95 (syn/anti 45:55)
4
5
6
7
3 equiv
3 equiv
3 equiv
3 equiv
16d
16e
16 f
16g
95 (syn/anti 45:55)
85 (syn/anti 45:55)
87
89
8
3 equiv
3 equiv
3 equiv
16h
16i
16j
92
9
85
10
85 (2 dia 50:50)
11
12
3 equiv
5 equiv
16k
77 (2 dia 55:45)
16l
49
16m
33 (syn/anti 45:55)
Chem. Eur. J. 2014, 20, 8918 – 8922
www.chemeurj.org
8921
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim

Communication
Table 6. (Continued)
Entry
13
E+
Equiv
Products
Yield [%][a]
5 equiv
16n
16 f
50 (2 dia 60:40)
40
14
5 equiv
16o
54 (2 dia 60:40)
[a] Yield of the isolated products based on starting material 1.
[5] For recent reviews on the chemistry of allenes, see: a) Modern Allene
Chemistry (Eds: N. Krause, A. S. K. Hashmi), Wiley-VCH, Weinheim, 2004;
b) S. Ma, Chem. Rev. 2005, 105, 2829. For some recent examples of car-
bometalation of allenes, see:c) S. Araki, H. Usui, M. Kato, Y. Butsugan, J.
Am. Chem. Soc. 1996, 118, 4699; d) T. Nishikawa, H. Shinokubo, K.
Oshima, Org. Lett. 2003, 5, 4623; e) Z. Lu, S. Ma, Adv. Synth. Catal. 2007,
349, 1225; f) Y. Yoshida, K. Murakami, H. Yorimitsu, K. Oshima, J. Am.
Chem. Soc. 2010, 132, 8878; g) S. Li, B. Miao, W. Yuan, S. Ma, Org. Lett.
2013, 15, 977.
late generation, application of this “protect and direct” strategy
to the ketonic carbonyl group should prove to be a very fruit-
ful area for future research.
Keywords: allene · carbometalation · copper · enolate ·
Grignard reagent
[6] a) A. Levin, A. Basheer, I. Marek, Synlett 2010, 329; b) Y. Minko, M. Pasco,
L. Lercher, M. Botoshansky, I. Marek, Nature 2012, 490, 522; c) Y. Minko,
M. Pasco, H. Chechik, I. Marek, Beilstein J. Org. Chem. 2013, 9, 526.
[7] A. Alexakis, J. F. Normant, J. Villieras, J. Organomet. Chem. 1975, 96, 471.
[1] For reviews, see: a) Preparation of Alkenes (Ed: J. M. J. Williams), Oxford
University Press, Oxford, 1996; b) A. B. Flynn, W. W. Ogilvie, Chem. Rev.
2007, 107, 4698.
[2] For a review see: J. F. Normant, A. Alexakis, Synthesis 1981, 841.
[3] For reviews, see inter alia: a) J. V. N. Vara Prasad, C. N. Pillai, J. Organo-
met. Chem. 1983, 259, 1; b) B. Breit, Y. Schmidt, Chem. Rev. 2008, 108,
2928; c) A. G. Fallis, P. Forgione, Tetrahedron 2001, 57, 5899.
ˇ
ˇ
[8] For preparation of the b-allenyl ester 1a, see: a) A. Duris, D. M. Barber,
H. J. Sanganee, D. J. Dixon, Chem. Commun. 2013, 49, 2777; b) K.
Kimura, S. Tanaka, Y. Tamaru, Bull. Chem. Soc. Jpn. 1995, 68, 1689.
[9] J. F. Normant, Pure Appl. Chem. 1978, 50, 709.
[4] For reviews, see: a) A. H. Hoveyda, D. A. Evans, G. C. Fu, Chem. Rev.
1993, 93, 1307; b) A. G. Fallis, P. Forgione, Tetrahedron 2001, 57, 5899;
c) A. Basheer, I. Marek, Beilstein J. Org. Chem. 2010, 6, 77; d) E. Erdik, Tet-
rahedron 1984, 40, 641. For recent examples of carbometalation of al-
kynes, see: e) X. Zhang, Z. Lu, C. Fu, S. Ma, Org. Biomol. Chem. 2009, 7,
3258; f) S. J. Aves, K. G. Pike, D. R. Spring, Synlett 2010, 2839; g) A.
Unsinn, C. Dunst, P. Knochel, Beilstein J. Org. Chem. 2012, 8, 2202.
[10] a) I. Salkind, T. Beburischwili, Ber. Dtsch. Chem. Ges. 1909, 42, 4500;
b) M. J. Jorgenson in Organic Reactions (Ed.: W. G. Dauben) 1970, 18, 1,
John Wiley & Sons.
Received: April 29, 2014
Published online on June 24, 2014
Chem. Eur. J. 2014, 20, 8918 – 8922
www.chemeurj.org
8922
ꢀ 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim