Cascade resulting in the reductive ethynylation of aldehydes: Dissection of its components

Article abstract of DOI:10.1021/ja910825g

A mild and efficient two-carbon homologation of aldehydes exploiting multiple modes of KOt-Bu was developed. This process involves a sequential Peterson allenylation/allene-alkyne isomerization/ protodesilylation in a single-flask operation. The differential roles played by the various elements of the process were demonstrated through dissection experiments.

Full text of DOI:10.1021/ja910825g

Published on Web 03/04/2010
Cascade Resulting in the Reductive Ethynylation of
Aldehydes: Dissection of Its Components
Dongjoo Lee† and Samuel J. Danishefsky*,†,‡
Laboratory for Bioorganic Chemistry, Sloan-Kettering Institute for Cancer Research,
1275 York AVenue, New York, New York 10065, and Department of Chemistry, Columbia
UniVersity, HaVemeyer Hall, 3000 Broadway, New York, New York 10027
Received December 23, 2009; E-mail: s-danishefsky@ski.mskcc.org
Abstract: A mild and efficient two-carbon homologation of aldehydes exploiting multiple modes of KOt-Bu
was developed. This process involves a sequential Peterson allenylation/allene-alkyne isomerization/
protodesilylation in a single-flask operation. The differential roles played by the various elements of the
process were demonstrated through dissection experiments.
Introduction
ethynylation of an aldehyde (i.e., conversion of aldehyde 5 to a
propargyl derivative 2) by nucleophilic addition of some reagent.
The Corey-Fuchs sequence1 elegantly accomplishes the
extension by one carbon of an aldehyde to a terminal
acetylene. It also anticipates further functionalization of the
alkyne, via the metal acetylide bond present in the concluding
phase of the process. In the research described below, we
posed a different question, as summarized in Scheme 1. We
wondered about the possibility of extending a chain (RCn,
see 1) by two carbon centers to a terminal ethynyl unit (see
target 2). In principle, there is a well-established method for
doing so via a leaving group at the terminus of the Cn moiety
(see 3), which serves as an electrophile for metalated
acetylene 4 in an alkylation format.2 However, all too often,
the formation of a carbon-carbon bond by alkylating a
metalated acetylide with a carbon electrophile bearing a
leaving group, even at a primary center, is not a straight-
forward matter, particularly when there is some steric
hindrance to attack at the alkylating carbon.3
We also set as a strong preference that the process be executable
by a very simple protocol. A cascade to accomplish this end
was designed and its components were dissected and individu-
ally demonstrated. As our true carbon nucleophile, we made
use of compound 7,5 which is readily obtained by the hydrozir-
conation of bis-trimethylsilylacetylene. The thought behind the
use of 7 was the expectation that carbon-based anion stabilizing
groups, such as TMS groups, might accelerate a required allene-
alkyne isomerization (vide infra).
Results and Discussion
A series of 1,2-bis(TMS)allylic alcohols (see Scheme 2 and
Table 1), were prepared via the reaction of 7 with the required
aldehydes. We next considered a proposed allenylation by
sequential Brook6 and Peterson7 transformations. It was found
that the use of n-BuLi, per se, did not accomplish the desired
allenylation. Apparently, the relatively covalent alkoxylithium
bond, formed from the deprotonation of the alcohol, is unable
to trigger the required Brook-Peterson reaction.8 We took note
that previous groups had investigated the possibility of using
catalytic amounts (ca. 10 mol % of KH) and R-TMS substituted
allylic alcohol, seeking to accomplish allenylation. Only low
yields of allene products were obtained.9-11 Apparently, under
It was our perception that addition of metalated acetylides,
such as 4, to aldehydes is more widely applicable than is “RX
alkylation.” Of course, addition of 4 to aldehyde 5 produces a
secondary propargylic alcohol (see 4 f 6). Progression from 6
to 2 would require site specific deoxygenation.4 What we were
seeking was implementable chemistry to enable overall reductiVe
† Sloan-Kettering Institute for Cancer Research.
‡ Columbia University.
(5) (a) Rosenthal, U.; Ohff, A.; Baumann, W.; Tillack, A. Z. Anorg. Allg.
Chem. 1995, 621, 77–83. (b) Sun, H.; Burlakov, V. V.; Spannenberg,
A.; Baumann, W.; Arndt, P.; Rosenthal, U. Organometallics 2002,
21, 3360–3366.
(1) Corey, E. J.; Fuchs, P. L. Tetrahedron Lett. 1972, 3769–3772.
(2) Selected examples: (a) Johnson, J. R.; Schwarz, A. M.; Jacobs, T. L.
J. Am. Chem. Soc. 1938, 60, 1882–1884. (b) Rutledge, T. F. J. Org.
Chem. 1959, 24, 840–842. (c) Kotsuki, H.; Kadota, I.; Ochi, M.
Tetrahedron Lett. 1990, 31, 4609–4612.
(6) (a) Brook, A. G. Acc. Chem. Res. 1974, 7, 77–84. (b) Brook, A. G.;
Bassindale, A. R. In Rearrangements in Ground and Excited States;
De Mayo, P., Ed.; Academic Press: New York, 1980; Vol. 2, pp 149-
227. (c) Colvin, E. W. In Silicon in Organic Synthesis; Butterworths:
London, 1981; pp 30-37.
(3) (a) Frye, L. L.; Robinson, C. H. J. Org. Chem. 1990, 55, 1579–1584.
(b) Anderson, J. C.; Denton, R. M.; Hickin, H. G.; Wilson, C.
Tetrahedron 2004, 60, 2327–2335. (c) Blizzard, T. A. et al. Bioorg.
Med. Chem. Lett. 2006, 16, 834–838. (d) Ko¨gl, A.; Brecker, L.;
Warrass, R.; Mulzer, J. Eur. J. Org. Chem. 2008, 2714–2730.
(4) Selected examples: (a) Hennion, G. F.; Sheehan, J. J. J. Am. Chem.
Soc. 1949, 71, 1964–1966. (b) McComsey, D. F.; Reitz, A. B.;
Maryanoff, C. A.; Maryanoff, B. E. Synth. Commun. 1986, 16, 1535–
1549. (c) Diaz, D.; Martin, V. S. Tetrahedron Lett. 2000, 41, 743–
746. (d) Trost, B. M.; Rudd, M. T. J. Am. Chem. Soc. 2005, 127,
4763–4776. (e) Nishibayashi, Y.; Shinoda, A.; Miyake, Y.; Matsuzawa,
H.; Mitsunobu, S. Angew. Chem., Int. Ed. 2006, 45, 4835–4839.
(7) (a) Ager, D. J. Org. React. 1990, 38, 1–223. (b) Kano, N.; Kawashima,
T. In Modern Carbonyl Olefination; Takeda, T., Ed.; Wiley-VCH:
Weinheim, Germany, 2004; pp 18-103.
(8) The bond dissociation energy of the Li-O bond (82 kcal/mol) is higher
than that of the K-O bond (57 kcal/mol) and the Na-O bond (61
kcal/mol). Lange’s Handbook of Chemistry, 15th ed.; Dean, J. A., Ed.;
McGraw-Hill: New York, 1999.
(9) Sato, F.; Tanaka, Y.; Sato, M. J. Chem. Soc., Chem. Commun. 1983,
165–166.
9
10.1021/ja910825g  2010 American Chemical Society
J. AM. CHEM. SOC. 2010, 132, 4427–4430 4427

A R T I C L E S
Lee and Danishefsky
Scheme 1
Scheme 2
Scheme 3. KOt-Bu Promoted Multiple Transformations
these conditions, the vinyl potassium species arising from the
Brook step is protonated by remaining free alcohol. Thus, allyl
silyl ethers were obtained from resultant intermediates with the
as-yet unreacted alcohol.
We hoped that this problem could be circumvented through
the use of n-BuLi as the base,12 thereby leading to high-yielding
lithium alkoxide formation (Scheme 3). This lithio alkoxide
would not undergo the Brook rearrangement required for
Peterson allenylation in a relatively low polarity solvent such
(10) Tius, M. A.; Pal, S. P. Tetrahedron Lett. 2001, 42, 2605–2608.
(11) Wilson, S. R.; Georgiadis, G. M. J. Org. Chem. 1983, 48, 4143–
4144.
(12) Tsubouchi, A.; Kira, T.; Takeda, T. Synlett 2006, 2577–2580.
9
4428 J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010

Reductive Ethynylation of Aldehydes
A R T I C L E S
Table 1. Scope of KOt-Bu Promoted One-Pot Alkynylation
solvent, a 67% yield of allenylsilane 13, accompanied by ca.
17% of the corresponding TMS acetylide 14, was obtained. We
next treated the isolated 13 with 0.3 equiv of KOt-Bu, thereby
obtaining an 83% yield of 14. Combining this information, we
returned to 12. When a solution of 12 in Et2O was treated with
1.3 equiv of n-BuLi and 1.3 equiv of KOt-Bu, an 85% yield of
14 was obtained in a single step. We note parenthetically that,
ordinarily, allene f acetylene isomerizations require the use
of very strong bases.17,18 That it was accomplished here simply
with t-butoxide reflects the predicted activating effect of the
terminal TMS group.
We then probed the possibility that nucleophile-induced
desilylation of the terminal acetylide could be accomplished by
KOt-Bu in a more polar solvent (such as HMPA). This
possibility was explored with substrate 14. Indeed, as shown,
the reaction thus conducted gave rise to 15 in 87% yield.
Combining these informatics, a one step protocol, leading from
12 to 15 in 81% yield, was implemented. The high yielding
cascade involves cooperativity between n-BuLi to achieve clean
deprotonation of 12, and transmetallation with KOt-Bu to
accomplish Brook rearrangement. This is followed by elimina-
tion, allene f acetylide isomerization enabled by the terminal
TMS group and exposure to the nucleophilicity enhancing
character of HMPA (to enable desilylation).
Informed by these dissection experiments, we investigated
the scope of the cascade type reaction. As shown in Table 1,
the reductive ethynylation chemistry was found to be compatible
with a range of alkyl-substituted allylic alcohols, obtained from
the reactions of the appropriate aldehydes with 7 (entries 1-6).
We were particularly pleased to observe that the sterically
hindered neo-pentyl alcohols, 26 and 29 (entries 5 and 6),
derived from their hindered aldehyde precursors, readily entered
into the cascade to provide the corresponding alkynes 27 and
30 in good yields.
We also investigated application of this sequence to a
conjugated enal (see 31). In this case, while the chemistry
described above worked, a three-component mixture of products
E,Z-33 and 34 was obtained. This result is understood in terms
of the ambident protonation possibilities of 35, generated in the
course of the allenyl silane f ethynylsilane isomerization. Also
not surprising was the finding that attempted application of the
reductive ethynylation protocol to 37 led to 38. The formation
of the latter is seen to reflect late-stage tautomerization of the
initially produced terminal alkyne, facilitated by the aryl group
of the primary product.
as Et2O.13,14 We hoped that transmetallation (Li f K),15 after
successful lithium alkoxide formation, could be readily ac-
complished and that the potassium salt of the allylic alcohol,
thus produced efficiently, would undergo high yielding Brook
rearrangement. This possibility was investigated, starting with
aldehyde 11, which reacted with 7 to give rise to 12 as shown.
Deprotonation of 12 with n-BuLi (1.3 equiv) was followed by
addition of 1.0 equiv of KOt-Bu.16 In Et2O as the primary
Finally, we sought to extend this chemistry to a multicom-
ponent setting, wherein the ultimate alkynyl anion would be
trapped by an electrophilic species (Scheme 4). We first
(17) (a) Thompson, H. W. J. Org. Chem. 1967, 32, 3712–3713. (b)
Baudouy, R.; Delbecq, F.; Gore, J. Tetrahedron Lett. 1979, 20, 937–
940. (c) Balme, G.; Doutheau, A.; Gore, J.; Malacria, M. Synthesis
1979, 508–510. (d) Michelot, D.; Clinet, J. C.; Linstrumelle, G. Synth.
Commun. 1982, 12, 739–747. (e) Zweifel, G.; Hahn, G. J. Org. Chem.
1984, 49, 4565–4567. (f) Hahn, G.; Zweifel, G. Synthesis 1983, 883–
885. (g) Michelot, D. Synth. Commun. 1989, 1705–1711. (h) Spence,
J. D.; Wyatt, J. K.; Bender, D. M.; Moss, D. K.; Nantz, M. H. J. Org.
Chem. 1996, 61, 4014–4021.
(13) For the effect of solvent polarity on silyl migration, see: Shinokubo,
H.; Miura, K.; Oshima, K.; Utimoto, K. Tetrahedron 1996, 62, 503–
514.
(18) Terminal allene (see below) didn’t undergo allene-alkyne isomerization
with 1.3 equiv of KOt-Bu in Et2O at ambient temperature.
(14) In the absence of KOt-Bu, the lithium alkoxide didn’t undergo
Brook-Peterson allenylation even at room temperature in either
Et2O or THF.
(15) For the Brook rearrangement with transmetallation of Li f K, see:
Smith, A. B., III; Xian, M.; Kim, W.-S.; Kim, D.-S. J. Am. Chem.
Soc. 2006, 128, 12368–12369.
(16) 1M solution of potassium tert-butoxide (KOt-Bu) in THF was
purchased from Aldrich Chemical Co.
9
J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010 4429

A R T I C L E S
Lee and Danishefsky
Scheme 4. KOt-Bu Promoted One-Pot Multicomponent Approach
flask sequence. The key components of the cascade include
Brook-Peterson allenylation, allene-alkyne isomerization, and
protodesilylation or other means of trapping the ultimate
potassium acetylide. Overall, the transformation results in two-
carbon homologation of aldehydes to provide the corresponding
terminal alkynes, or other trapping products. The differential
roles played by the various elements of the process were
demonstrated through dissection experiments. Further studies
of the scope and application of these findings are in progress.
Acknowledgment. Support for this research was provided by
the National Institutes of Health (CA103823 to SJD). We thank
Ms. Rebecca Wilson for editorial consultation and Ms. Dana Ryan
for assistance with the preparation of the manuscript. We thank
Dr. George Sukenick, Ms. Hui Fang, Ms. Sylvi Rusli, and Ms. Hui
Liu of the Sloan-Kettering Institute’s NMR core facility for mass
spectral and NMR spectroscopic analysis.
examined this concept in the context of a deuterium oxide
quench. Thus, sequential treatment of the lithium alkoxide
generated from alcohol 12 with KOt-Bu (3 equiv), HMPA (5
equiv), and finally D2O, provided the deuterated alkyne 3919 in
74% overall yield. This type of quenching experiment was found
to be feasible with other electrophiles. For instance, trapping
by either MeI or, indeed, by aldehyde 25, led to the formation
of 40 and 41, respectively, as shown.
Supporting Information Available: General experimental
procedures, including spectroscopic and analytical data for new
compounds along with copies of 1H and 13C NMR spectra. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Conclusion
In conclusion, we have developed a mild and efficient process
which exploits multiple reactivity modes within a single one-
(19) Deuterium incorporation was greater than 95%.
JA910825G
9
4430 J. AM. CHEM. SOC. VOL. 132, NO. 12, 2010