Synthesis of enantiopure tertiary skipped diynes via one-pot desymmetrizing TMS-cleavage

Article abstract of DOI:10.1021/ol3017462

Enantiopure tetrasubstituted skipped diynes were readily synthesized from N-protected amino esters upon addition of lithium TMS-acetylide which was found to be desymmetrizing through one-pot selective TMS-cleavage. The deprotection of the TMS group was realized through a one-pot silicon atom attack by the liberated methoxide, which was diastereoselective due to a conformational favorable chelate.

Full text of DOI:10.1021/ol3017462

ORGANIC
LETTERS
2012
Vol. 14, No. 15
3974–3977
Synthesis of Enantiopure Tertiary Skipped
Diynes via One-Pot Desymmetrizing
TMS-Cleavage
Malek Nechab*,† and Nicolas Vanthuyne‡
ꢀ
Aix-Marseille Universite, CNRS, ICR, UMR 7273, 13397 Cedex 20, Marseille, France,
and Aix-Marseille Universite, CNRS, ISM2, 13397 Cedex 20, Marseille, France
ꢀ
malek.nechab@univ-amu.fr
Received June 26, 2012
ABSTRACT
Enantiopure tetrasubstituted skipped diynes were readily synthesized from N-protected amino esters upon addition of lithium TMS-acetylide
which was found to be desymmetrizing through one-pot selective TMS-cleavage. The deprotection of the TMS group was realized through a
one-pot silicon atom attack by the liberated methoxide, which was diastereoselective due to a conformational favorable chelate.
Skipped diynes are 1,4-diyn-1,5-diyl units where an sp3
carbon separates the two triple bounds.1,2 They are valu-
able synthons for the construction of carbon-rich struc-
tures and precursors of 1,4-cis,cis-dienes present in many
bioactive compounds.3 Tertiary chiral diynes are impor-
tant building blocks that could be used in a broad array
of reactivities and transformed into more complex
compounds.4 Diederich has proven the utility of 1,4-diynes
as precursors of chiral oligomers,5 two- and three-
dimensional alleno-acetylenic chiral macrocycles.6
Chauvin and co-workers have described the synthesis of
carbo-benzene from 1,4-dialkynylbutatrienes.7 Another
elegant application is the enantioselective synthesis of
cyclopentenones via rhodium-catalyzed desymmetrization
of 4-diynals reported by Fu.8
Recently, skipped diynes were also used in metal-cata-
lyzed intramolecular cyclization to reach heterocycles9
such as pyrrolines.10
Very recently, our group has been involved in the
development of a cascade rearrangement of enediynes.11
In continuation of these investigations, we were interested
(7) (a) Maraval, V.; Leroyer, L.; Harano, A.; Barthes, C.; Saquet, A.;
Duhayon, C.; Shinmyozu, T.; Chauvin, R. Chem.;Eur. J. 2011, 17,
5086–5100. For a review, see: Marval, V.; Chauvin, R. Chem. Rev. 2006,
106, 5317–5343.
†
ꢀ
ꢀ
Aix-Marseille Universite, CNRS, ICR.
‡
Aix-Marseille Universite, CNRS, ISM2.
(1) Tedeschi, C.; Saccavini, C.; Maurette, L.; Soleilhavoup, M.;
Chauvin, R. J. Organomet. Chem. 2003, 670, 151–169.
(8) Tanaka, K.; Fu, G. C. J. Am. Chem. Soc. 2002, 124, 10296–10297.
ꢀ
ꢀ
(2) Tejedor, D; Lopez-Tosco, S; Gonzalez-Platas, J.; Garcıa-Tellado,
ꢀ
€
(9) Ruttinger, R.; Leutzow, J.; Wilsdorf, M.; Wilckens, K.; Czekelius,
F. J. Org. Chem. 2007, 72, 5454–5456.
(3) Durand, S.; Parrain, J.-L.; Santelli, M. J. Chem. Soc., Perkin
Trans. 1 2000, 253–273.
C. Org. Lett. 2011, 13, 224–227 and references cited therein.
(10) Bhaskara, V.; Iska, R.; Verdolino, V.; Wiest, O.; Helquist, P.
J. Org. Chem. 2010, 75, 1325–1328.
ꢀ
ꢀ
ꢀ
(4) (a) Tejedor, D.; Lopez-Tosco, S.; Gonzalez-Platas, J.; Garcıa-
ꢀ
(11) (a) Nechab, M.; Campolo, D.; Maury, J.; Perfetti, P.;
Vanthuyne, N.; Siri, D.; Bertrand, M. P. J. Am. Chem. Soc. 2010, 132,
14742–14744. (b) Nechab, M.; Besson, E.; Campolo, D.; Perfetti, P.;
Vanthuyne, N.; Bloch, E.; Denoyel, R.; Bertrand, M. P. Chem. Commun.
2011, 47, 5286–5288. (c) Campolo, D.; Gaudel-Siri, A.; Mondal, S.; Siri,
D.; Besson, E.; Vanthuyne, N.; Nechab, M.; Bertrand, M. P. J. Org.
Chem. 2012, 77, 2773–2783. (d) Mondal, S.; Nechab, M.; Vanthuyne, N.;
Bertrand, M. P. Chem. Commun. 2012, 48, 2549–2551. (e) Mondal, S.;
Nechab, M.; Campolo, D.; Vanthuyne, N.; Bertrand, M. P. Adv. Synth.
Catal. 2012, 354, 1987–2000.
Tellado, F. Chem.;Eur. J. 2011, 17, 9571–9575. (b) Tejedor, D.; Lopez-
ꢀ
ꢀ
Tosco, S.; Gonzalez-Platas, J.; Garcıa-Tellado, F. Chem.;Eur. J. 2010,
16, 3276–3280.
(5) (a) Ter Wiel, M. K. J.; Odermatt, S.; Schanen, P.; Seiler, P.;
Diederich, F. Eur. J. Org. Chem. 2007, 3449–3462. (b) Buschhaus, B.;
Convertino, V.; Rivera-Fuentes, P.; Alonso-Gomez, J. L.; Petrovic,
ꢀ
A. G.; Diederich, D. Eur. J. Org. Chem. 2010, 2452–2456.
ꢀ
(6) Alonso-Gomez, J. L.; Rivera-Fuentes, P.; Harada, N.; Berova,
N.; Diederich, F. Angew. Chem., Int. Ed. 2009, 48, 5545–5548.
r
10.1021/ol3017462
Published on Web 07/19/2012
2012 American Chemical Society

in getting access to pyrrolidines from skipped diynes via
the copper-carbenoid12 mediated allenoate formation fol-
lowed by an intramolecular aminocyclization according
to a methodology developed by Tan.13 During attempts to
reach the substrates, an unprecedented phenomenon was
evidenced. Results are detailed hereafter.
it with a methyl group, i.e. Ts-Me-Ala-OMe (entry 6,
Table 1), only furnished 1c in 61% isolated yield, and
TMS-cleavage did not occur. This result suggests that the
amide anion is presumably involved in a chelated species.
Table 1. Addition of TMS-Acetylene on Amino Esters (AE)
Leading to 1 or 2
Scheme 1. Addition of TMS-Acetylene on Amino Esters Leading
to 1 or 2
entry
AEa
Glyd
t (h)
1 (%)b
2 (%)b
drc
1
4
a
b
b
b
b
c
d
e
f
27.5
nde
nd
14
8
37
71
ꢀ
2
Ala
4
5:1
5:1
ꢀ
3
(S)-Ala
Ala
4
70 (ee >98)
4
4f
4g
4h
4i
5
33
5
Ala
58
5:1
ꢀ
6
Ala
61
19
nd
nd
82
82
83
nd
nd
34
57
ꢀ
7
Ala
55
6:1
9:1
4:1
ꢀ
8
Phg
74
9
Met
5
69
As an initial target, we started with the synthesis of the
diyne 1a in a classical way (Scheme 1 and entry 1, Table 1).
In the first attempt, to 4 equiv of trimethylsilylacetylene
and n-BuLi in THF at ꢀ78 °C was added Ts-glycine ethyl
ester. Unexpectedly, it was observed that, by warming the
reaction to rt, the product 2a was obtained in 37% yield
together with only 27% of the expected product 1a. A
desymmetrizationofdiyne2aby a selective deprotection of
the TMS group was induced by warming the reaction. This
resultencouragedus tostudyfurther the desymmetrization
process because this class of compounds is offering a large
potential of applications after their postfunctionalization.
Interestingly, when Ts-alanine methyl ester (Ts-Ala-
OMe) was submitted to similar experimental conditions,
the monoprotected diyne 2b was isolated as the major
product in 71% yield. More interestingly, the diastereo-
selective cleavage of one trimethylsilyl group led to a diaster-
eomeric ratio (dr) of 5:1.14 Furthermore, despite the use of
an excess of n-BuLi, no racemization occurred when an
enantiopure amino ester was used; the enantiomeric excess
of the tetrasubstituted diyne was higher than 98% (entry 3,
Table 1). The procedure was realized in gram scale (2.6 g of
substrate).15 It is also important to mention that the dr of
2b is constant during the reaction time.
10
11
12
13
14
15
16
Lactate
pGlu
Pro
4j
24k
4
g
h
i
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
(S)-Trp
Ser
4
j
65 (ee >99)
4:1
4:1
2:1
ꢀ
4
k
l
50
38
ꢀ
Val
5
β-Ala
4
m
a (()-N-Ts methyl amino ester elsewhere mentioned. b Isolated yield.
c Diastereomeric ratio determined by 1H NMR of the crude mixture.
d Ts-Gly-OEt was used. e Not determined, yield <5%. f 3 equiv of
acetylide were used. g MeLi was used instead of n-BuLi. h Ts-Me-Ala
was used instead of (()-Ts-Ala. i Ac-Ala was used instead of Ts-Ala.
j (()-Methyl lactate was used instead of Ts-Ala. k (()-Methyl pyrroglu-
tamate was used.
At this stage of the investigation, rationalizing the
mechanism of the TMS-removal is not obvious.16 Cleavage
with butyl lithium has never been reported, but MeLi-
induced deprotection is known to be possible.17 When the
reaction was performed at ꢀ78 °C, the expected diyne 1b
was the sole product which confirms that 1b is the inter-
mediate in the formation of 2b. In order to elucidate the role
of n-BuLi in the desymmetrization, we subjected diyne 1b to
a reaction with 1, 2, or 3 equiv of n-BuLi at room tempera-
ture (Scheme 2). After 4 h, only 15% of 2b were formed and
1b was recovered. We can then conclude that n-BuLi alone is
not implicated in the cleavage of the TMS group.
A second mechanism is possible: Another nucleophile
which is present in the medium (namely MeOꢀ generated
from the alanine methyl ester) could be responsible
for the silicon atom attack. To validate this hypothesis,
the diyne 1b was subjected to the reaction with lithium
methoxide (4 equiv) in THF to reproduce nearly the same
conditions as the one-pot desymmetrization process
(Scheme 2). After 4 h, the diyne 1b was completely
consumed and 2b was formed in 46% yield. The diyne 2b
In order to further explore the system, we envisaged
different experiments. The addition of 3 equiv of lithium
acetylide to the amino ester afforded 33% of the desired
product 2b and 14% of 1b (entry 4). 22% of the free diyne
(double deprotection) was also isolated which was also
enantiopure. When the reaction was conducted with TMS-
acetylide generated with MeLi (entry 5), only a small
difference in yield was observed as compared to n-BuLi
conditions. Suppressing the amide hydrogen by replacing
(12) For an example of furylcyclobutene synthesis from diazoace-
ꢀ
tates and diynes, see: Barluenga, J.; Riesgo, L.; Lopez, L. A.; Rubio, E.;
ꢀ
Tomas, M. Angew. Chem., Int. Ed. 2009, 48, 7569–7572.
(13) Liu, H.; Feng, W.; Kee, C. W.; Leow, D.; Loh, W.-T.; Tan, C.-H.
(16) For the deprotection of the TMS group, see: (a) Greene, T. W.;
Wuts, P. G. M. Protective Groups in Organic Synthesis, 4th ed.; Wiley and
Sons: New York, 2000. (c) For review, see: Crouch, R. D. Tetrahedron 2004,
60, 5833. (d) For a selective TMS acetylene group cleavage in the
presence of alkyl silyl ethers, see: Yeom, C.-E.; Kim, M. J.; Choi,
W. B.; Kim, M. Synlett 2008, 565–568.
Adv. Synth. Catal. 2010, 352, 3373–3379.
(14) For asymetric synthesis of diynes from optically active propargyl
ketone, see: Convertino, V.; Manini, P.; Schweizer, W. B.; Diederich, F.
Org. Biomol. Chem. 2006, 4, 1206–1208.
(15) When a preliminary generated Ts-Ala-OMe amide anion was
added to 3 equiv of lithium acetylide, the same yield was observed.
(17) Duhamel, P.; Cahard, D.; Quesnel, Y.; Poirier, J.-M. J. Org.
Chem. 1996, 61, 2232–2235.
Org. Lett., Vol. 14, No. 15, 2012
3975

had a diastereomeric ratio identical to that observed
above. Additionally, 29% of free diyne (removal of the
two TMS) werealsoisolated. Inlight of these observations,
we can postulate that a chelated lithium methoxide, formed
during the addition of the acetylide to the amino ester, is most
probably responsible for the TMS-cleavage. Moreover, when
the addition of lithium acetylide was carried out on Ts-Ala-
O-t-Bu instead of Ts-Ala-OMe, the ratio of 2b:1b was only
1:4. This result confirms the chelation hypothesis in that the
bulky alkoxide (t-BuOꢀ) would be less effective in the attack
of the TMS group.
An analogy could be made between the chelation of
MeOLi with compound 1 and TMEDA-BuLi chelation. A
monomer- or dimer-based pathway could be involved.18
The pro-(S) attack of the chelated methoxide on the silicon
atom would be diastereoselective giving rise to (S,S)-2.
This process is probably sequential (with preliminary
methoxide dissociation) and not concerted because of a
nonfavored constrained transition state of the attack. The
pro-(R) attack would be disfavored by the tosyl group; this
face is hindered as illustrated by the simplified Chem-3D
model of the monomer (Figure 1).19 Therefore, the sulfo-
nyl group could play a role in driving the MeOLi nucleo-
phile toward the pro-(S) trimethylsilylethynyl substituent.
Another possible explanation would be the formation
of a chiral trans-dimer or any analogue (Scheme 3). This
dimer is C2-symmetric and could be predominant in the
medium,20 by way of favoring the formation of the major
(S,S)-diastereomer. The suggested relative stereochemistry
was confirmed by an NOE sequence after derivatization of
2 (vide infra).
Scheme 2. Role of n-BuLi and MeOLi in the TMS-Removal
A more complex tetramer aggregate could also be
hypothesized since these complexes have already been
discussed.21
These hypotheses are coherent; the reaction of Ts-Val-
OMe showed low diastereoselectivity because of a bulky
isopropyl group which does not allow differentiation of
the two faces in the TMS deprotection. This was also
confirmed by a longer reaction time in the formation of
2l as compared to 2b (entry 15, Table 1).
The mechanism involving a five-membered chelate
would be preferred since the use of a β-alanine ester
substrate (entry 16, Table 1) did not afford deprotection
of the TMS group, and only compound 1m was isolated.
This is again in agreement with the proposed chelate.
In order to demonstrate the relative stereochemistry,
we coupled diynes 2 to ethyl diazoacetate in the presence
of a catalytic amount of copper(I) iodide22 to prepare the
pyrrolidine motif (Scheme 4).23
The reaction scope of different amino esters has been
then studied in the presence of 4 equiv of lithium acetylide.
Replacing the sulfonyl group by an acetyl one resulted in
slowing down the reaction rate. The yield of 2d was 55%
and 19% for 1d, but almost the same diastereoselectivity
was observed (entry 7, Table1).
N-Ts-phenylglycine methyl ester (Ts-Phg-OMe, entry 8)
gave similar results to those obtained with the alanine
analogue with the formation of 2e in 74% isolated yield
and 9:1 dr. A methionine-derived substrate was also tested
under the same conditions, and 2f was isolated in 69% with
drof 4:1 (entry 9). To better understand the mechanism, we
compared the reactivity of alanine and a lactate derivative.
It is interesting to mention that with the latter we observed
only the formation of 1g in 82% isolated yield and no trace
of 2g was detected (entry 10, Table 1). This behavior
indicates that the nitrogen atom is probably playing a
major role in the selective cleavage of the TMS group.
Lithium ligation by an amide anion should be involved in
the process as stated above.
When the diyne 2b was reacted with ethyl diazoacetate
in the presence of 5 mol % of CuI in acetonitrile, the
desired pyrrolidine 6 was obtained in 42% yield as a single
diastereoisomer with an ee >99%. Attempts to demon-
strate the relative stereochemistry of pyrrolidines 6 by
an NOE sequence were unsuccessful. We therefore
When methyl pyrroglutamate was used as the substrate
(entry 11), diyne 1h was formed in 82% yield. Reaction of
proline methyl ester showed the same behavior; i.e., no
TMS-cleavage occurred (entry 12). In the case of cyclic
substrates, the lack of flexibility or a constrained bicyclic
transition state would prevent the cleavage of the tri-
methylsilyl group.
The methodology was also applied to a tryptophan
derivative with free nitrogen at the indole moiety at gram
scale (3 g of substrate). Interestingly, the product 2i was
isolated in 65% yield with a 4:1 diastereomeric ratio and
high ee (>99.5%, entry 13). An unprotected serine analo-
gue gave also the same reactivity with lithium acetylide,
and the desymmetrized compound 2k was obtained in
satisfactory yield (50%) with a dr of 4:1 (entry 14).
(18) (a) Chadwick, S. T.; Rennels, R. A.; Rutherford, J. L.; Collum,
D. B. J. Am. Chem. Soc. 2000, 122, 8640–8647. (b) Qu, B.; Collum, D. B.
J. Am. Chem. Soc. 2005, 127, 10820–10821. (c) For a computational
study of lithium methoxide mixed aggregates, see: Pratt, L. M.; Kwon,
O.; Ho, T. C.; Nguyen, N. V. Tetrahedron 2008, 64, 5314–5321.
(19) (a) For a lithium dianion, see: Gruver, J. M.; West, S. P.; Collum,
D. B.; Sarpong, R. J. Am. Chem. Soc. 2010, 132, 13212–13213. (b) For
aminoalkoxide lithium aggregates, see: Khartabil, H. K.; Gros, P. C.;
ꢀ
Fort, Y.; Ruiz-Lopez, M. F. J. Am. Chem. Soc. 2010, 132, 2410–2416.
(c) For a review, see: Gros, P.; Fort, Y. Eur. J. Org. Chem. 2002, 3375–
3383.
(20) For an example of C2-symmetric Li-dimer, see: Martin, A.;
Kocks, B. M.; Spek, A. L.; Van Koten, G. J. Organomet. Chem. 2001,
624, 271–286.
(21) Jones, A. C.; Sanders, A. W.; Bevan, M. J.; Reich, H. J. J. Am.
Chem. Soc. 2007, 129, 3492–3493 and references cited therein.
(22) Suarez, A.; Fu, G. C. Angew. Chem., Int. Ed. 2004, 43, 3580–
3582.
(23) For details, see Supporting Information.
3976
Org. Lett., Vol. 14, No. 15, 2012

Scheme 3. Tentative Rationale for the Observed Diastereoselective TMS-Removal: Methoxide Attack after Its Dissociationa
a THF and Liþ have been removed for clarity.
Scheme 4. NOE Demonstration of the Relative Stereochemistry
Figure 1. Chem-3D model of the monomer chelate intermediate:
methoxide attack after its dissociation.
hydrogenated the alkyne moiety. The saturated derivative
which revealed a clear correlation between the methylene
hydrogens and the methyl group of 7 confirmed the
speculated relative stereochemistry and therefore (S,S)-
configurations.24
complex molecules is planned as well as theoretical calcu-
lations and will be reported in the future.
Acknowledgment. We thank ANR-10-JCJC-0701-01,
CNRS, and Aix-Marseille University for financial sup-
ꢁ
In summary, unexpected desymmetrizing deprotection
of a TMS group occurred upon addition of TMS-acetylide
on amino esters, leading to enantiopure tetrasubstituted
diynes. The latter were obtained in good diastereoselectiv-
ities and excellent ee’s. They represent an interesting
class of compounds which offer great potential in post-
functionalization. A preliminary sequence to access chiral
pyrrolidines has already been realized, and designing more
port. We are also grateful to Prof. Michele P. Bertrand
(Marseille), Dr. Jacques Einhorn (Grenoble), and Dr.
Benjamin Darses (ICSN) for helpful discussions.
Supporting Information Available. Experimental pro-
cedures and full characterization of new compounds.
This material is available free of charge via the Internet
at http://pubs.acs.org.
(24) (S,S)-Pyrrolidine derived from 2e was similarly assigned.
The authors declare no competing financial interest.
Org. Lett., Vol. 14, No. 15, 2012
3977