Fluoride-triggered domino reactions involving ammonium acetylides and carbonyl compounds

Article abstract of DOI:10.1002/ejoc.200901082

We describe the use of tetrabutylammonium fluoride as a basic trigger for reactions capable of generating structurally diverse products from methyl propiolate and carbonyl derivatives. The processes are based on either chemodifferentiating multicomponent ABB' three-component reactions or bimolecular domino reactions, and they operate through three different and well-defined autocatalytic cycles. These catalytic cycles share a common property: they are launched by the acid-base reaction of fluoride ions to give catalytic amounts of acetylide or enolate salts, but they are maintained by the autocatalytic generation of these salts.

Full text of DOI:10.1002/ejoc.200901082

SHORT COMMUNICATION
DOI: 10.1002/ejoc.200901082
Fluoride-Triggered Domino Reactions Involving Ammonium Acetylides and
Carbonyl Compounds
David Tejedor,*[a,b] Sara López-Tosco,[a,b] Gabriela Méndez-Abt,[a,b] and
Fernando García-Tellado*[a,b]
Keywords: Autocatalysis / Fluorides / Multicomponent reactions / Domino reactions / Alkynes
We describe the use of tetrabutylammonium fluoride as a ba-
sic trigger for reactions capable of generating structurally di-
verse products from methyl propiolate and carbonyl deriva-
tives. The processes are based on either chemodifferentiating
multicomponent ABBЈ three-component reactions or bimo-
lecular domino reactions, and they operate through three dif-
ferent and well-defined autocatalytic cycles. These catalytic
cycles share a common property: they are launched by the
acid–base reaction of fluoride ions to give catalytic amounts
of acetylide or enolate salts, but they are maintained by the
autocatalytic generation of these salts.
piolate to generate corresponding β-onium allenolate inter-
mediate I (Scheme 1), which is basic enough to deprotonate
either a molecule of alkyl propiolate to afford the corre-
sponding acetylide salt II (general for aldehydes and
ketones), or a molecule of the reactant carbonyl compound
to give enolate salt III (specific for unbranched 1,2-keto es-
ters).[5] Specific and skeletally different products are ob-
tained from these two salts through well-defined acetylide-
driven (structures 1–3)[4] or enolate-driven (isotetronic acid
derivative 5)[4b,5] domino processes. Whereas acid chlorides
afford skipped diynes 4 through a similar A2BBЈ four-com-
ponent reaction process,[6] aromatic 1,2-diketones them-
selves constitute a particular example of this reactivity gen-
eration concept, affording cyclobutanes 7 and cyclobutenes
8 through a well-defined domino reaction network involv-
Introduction
A recent communication from the Reboul group[1] de-
scribing the unprecedented in situ formation of cesium ace-
tylides of alkyl propiolates by catalytic amounts of cesium
fluoride encouraged us to report our own results dealing
with the catalytic formation of ammonium acetylides by
using tetrabutylammonium fluoride (TBAF) as the fluoride
source. As part of a wide research program aimed at the
design and development of new chemodifferentiating ABBЈ
three-component reactions (3CRs)[2] and higher homolo-
gous multicomponent reactions, we were interested in the
use of TBAF as a basic trigger for processes that can gener-
ate structurally diverse products from alkyl propiolates and
carbonyl derivatives. We chose this salt because, in aprotic
solvents, it offers a convenient balance between nucleophi-
licity (low) and base strength (between trialkylamines and
alkoxides).[3] We have elsewhere described[4] that these pro-
cesses can be conveniently triggered by tertiary amines and/
or tertiary phosphanes through a novel reactivity genera-
tion concept: the in situ generation of a strong base by a
good nucleophile. The concept is chemically expressed by
the Michael addition of the nucleophile on the alkyl pro-
ing ambiphilic allenes
6
as discrete intermediates
(Scheme 1).[7]
In this communication, we report on our preliminary fin-
dings in the use of naked fluoride ions dissolved in an apro-
tic medium as a convenient trigger for the chemodifferenti-
ation of ABBЈ 3CRs and related domino processes.
Results and Discussion
[a] Departamento de Química Biológica y Biotecnología,
Instituto de Productos Naturales y Agrobiología
Consejo Superior de Investigaciones Científicas
Astrofísico Francisco Sánchez 3
Scheme 2 outlines the chemical outcomes of the fluoride-
catalyzed reactions of methyl propiolate (alkynoate source)
and a set of carbonyl reactants spanning a wide and conve-
niently contrasted reactivity profile. Reactions were per-
formed in dichloromethane at room temperature and under
an aerobic atmosphere by mixing the two reactants in the
presence of catalytic amounts of the ion fluoride source
(10 mol-%, 1  in THF). The stoichiometry was adjusted
to that required for optimal product formation under atom-
38206 La Laguna, Tenerife, Spain
Fax: +34-922260135
E-mail: fgarcia@ipna.csic.es
dtejedor@ipna.csic.es
[b] Instituto Canario de Investigación del Cáncer
Canary Islands, Spain
www.icic.es
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.200901082.
Eur. J. Org. Chem. 2010, 33–37
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
33

D. Tejedor, S. López-Tosco, G. Méndez-Abt, F. García-Tellado
SHORT COMMUNICATION
Scheme 1. Lewis base catalyzed acetylide-driven and enolate-driven reactions.
economical conditions. Under these experimental condi- diketones react with methyl propiolate through acetylide-
tions, we were pleased to observe that benzil was stereose- driven processes, unbranched 1,2-keto esters react through
lectively transformed into a mixture of highly function- the enolate-driven pathway.
alized cyclobutane 9 and cyclobutane 10 in a modest com-
The three catalytic cycles share a common property: they
bined yield (40%). Whereas propionaldehyde and ethyl 3- are launched by the acid–base reaction of fluoride ions with
methyl-2-oxobutyrate afforded the expected 1,3-dioxolane propiolate or pyruvate to give catalytic amounts of enolate
derivatives 11 and 12 in excellent yields (89 and 91%, or acetylide salts, but these processes are maintained by the
respectively), ethyl pyruvate gave the corresponding homo- autocatalytic generation of these salts.[4] This property is
aldolic adduct 13 in modest yield (34%).[8] Remarkably, an chemically expressed in the form of an acid–base reaction
increase in the amount of fluoride ion from 5 to 25 mol-% between the most advanced allenolate intermediate (D, H,
(with respect to the carbonyl compound) delivered isote- or K) and the starting material to afford the product in the
tronic acid derivative 14 in 64% yield. Note that compound neutral final state and enolate (or acetylide) salt to reinitiate
14 incorporates a second unit of alkyl propiolate in the the catalytic cycle. Note that in these processes, the only
form of an enol ether functionality. In contrast, 1,2-keto- hydrogen source is the carbonyl compound itself or methyl
amides, which are transformed into the corresponding pro- propiolate.
pargylic derivative 1 under triethylamine catalysis,[4b] gener-
ated propargyl alcohol 15. Benzoyl chloride did not afford is more acidic than methyl propiolate. This is the case of
the corresponding 1,3-diyne 4 (R1 = Ph; R = Me).
ethyl pyruvate, which in the presence of ion fluoride forms
Cycle a (Scheme 3) is followed when the carbonyl partner
Scheme 3 outlines a mechanistic proposal accounting for the corresponding enolate A, which reacts with another
these experimental results. As was expected, when methyl molecule of ethyl pyruvate to afford the homoaldol adduct,
propiolate and the carbonyl compound are mixed in the which lactonizes to give the corresponding lactone B and
presence of TBAF, an acid–base reaction between fluoride ammonium ethoxide salt. An acid–base equilibrium involv-
and the more acidic species present in the reaction medium ing these species and enolate C is then established. If the
takes place. This acid–base reaction generates either enolate amount of fluoride is low (5 mol-%), enolate salt A, lactone
salt A or acetylide salt E and hydrogen fluoride. The pKa B, enolate C, ammonium ethoxide, and ethanol coexist in
balance between the alkyl propiolate and the carbonyl part- equilibrium. Enolate C is a mild nucleophile and it needs a
ner orchestrates these processes, affording the expected high concentration to productively add to methyl propiolate
products from either the enolate-driven cycle a or the ace- to keep the cycle going. If the amount of fluoride is in-
tylide-driven cycles b or c. Whereas aliphatic aldehydes, creased, the concentration of enolate C increases and the
branched 1,2-keto esters, 1,2-ketoamides, and aromatic 1,2- Michael addition of this enolate to the starting alkynoate
34
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 33–37

Fluoride-Triggered Domino Reactions
Scheme 2. Fluoride-catalyzed reactions involving methyl propiolate and carbonyl compounds.
Scheme 3. Fluoride-triggered domino process involving autocatalytic generation of acetylide or enolate salts.
Eur. J. Org. Chem. 2010, 33–37
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
35

D. Tejedor, S. López-Tosco, G. Méndez-Abt, F. García-Tellado
SHORT COMMUNICATION
shifts the equilibrium towards the formation of allenolate D.
In this step, HF could activate the alkynoate through H-
bond interactions.[3] Protonation of allenolate D affords iso-
tetronic acid derivative 14 and enolate salt A to reinitiate the
cycle (autocatalysis). Isotetronic acid derivative 14 is ob-
tained as a mixture of E/Z isomers, which is the expected
outcome from the Michael addition of alkoxide ions to alkyl
propiolates.[9] Because ethyl pyruvate is incorporated into
product 14 in the form of two chemodifferentiated structural
motives, we categorize this reaction as an ABBЈ 3CR.[2]
Cycle b (Scheme 3) operates through acetylide salt for-
mation. Obviously, it requires that the carbonyl partner be
less acidic than methyl propiolate. Once a catalytic amount
of acetylide salt E forms, it adds to the carbonyl compound
to generate propargylic alkoxide F, which in turn reacts with
a second unit of the carbonyl compound to give intermedi-
ate G. Intramolecular Michael addition yields allenolate H,
which in turn reacts with methyl propiolate to afford 1,3-
dioxolane derivatives 11 (or 12) and acetylide salt E to reini-
tiate the cycle (autocatalysis). 1,3-Dioxolane derivatives 11
and 12 are obtained as an isomeric mixture of the four pos-
sible isomers (syn/anti, E/Z) through a common chemodif-
ferentiated ABBЈ 3CR reaction. We can point out that the
isomeric mixtures of 1,3-dioxolanes were conveniently con-
verted into tetronic acid derivatives.[10] 1,2-Ketoamides con-
stitute a particular case of activated carbonyl compounds.
They react through cycle b (Scheme 3) to give the corre-
sponding quaternary alkoxide intermediate F (R1 = Me; R2
= CONBn2), which is a much stronger base than nucleo-
phile and it deprotonates the starting propiolate to reinitiate
the cycle. In this case, cycle b (Scheme 3) affords propar-
gylic amide 15, which incorporates just one unit of pro-
piolate and one unit of 1,2-ketoamide and it cannot be con-
sidered a domino process. Cycle c (Scheme 3) is particular
for aromatic 1,2-diketones and it also operates by acetylide
salt formation. The catalytic cycle involves 1,2-addition of
acetylide to the 1,2-diketone to form propargylic alkoxide
I, which rearranges into cumulenolate K. Acid–base proton-
ation of this intermediate generates ambiphilic allene L and
a new unit of acetylide salt E to reinitiate the cycle. Allene
L slowly dimerizes to give a mixture of C4-carbocycles 9
and 10. Overall, the reaction constitutes an impressive ex-
ample of a complexity-generating domino process.
Conclusions
In summary, we have described our results on the use of
TBAF as a basic trigger for domino processes involving
methyl propiolate and carbonyl derivatives. Although the
efficiency of these processes is sometimes lower than that
described for the Lewis base catalyzed versions,[4–7] fluoride
catalysis offers some advantages: (1) reactions are per-
formed at room temperature and under aerobic atmo-
spheres (bench-economy); (2) THF solutions of TBAF are
easily handled without especial care (CsF is a highly hygro-
scopic solid); (3) fluoride accumulates in the form of HF
(weak acid), which can participate as an H-bond donor[3]
(H-bond catalysis); (4) TBAF in aprotic solvents behaves as
a bad nucleophile and it is not expected to compete with
other nucleophiles for the electrophilic intermediates gener-
ated in these domino processes (i.e., allenes) (feasibility of
novel multicomponent reaction); (5) the results are comple-
mentary to those obtained when a much stronger base such
as nBuLi is utilized;[4a] (6) the ammonium counterion ac-
tively participates in each one of the three catalytic cycles
and it can be conveniently utilized as a chiral transfer cata-
lyst (ion-pair-based chiral implementation).[11]
Experimental Section
Compounds 9–14 have been fully described elsewhere.[4–7]
Fluoride-Catalyzed Reaction of N,N-Dibenzyl-2-oxopropanamide
and Methyl Propiolate: TBAF (1.0  in THF, 0.080 mmol) was
added to a solution of methyl propiolate (0.40 mmol) and N,N-
dibenzyl-2-oxopropanamide (0.40 mmol) in dry CH2Cl2 (5 mL).
The reaction mixture was stirred for 1 h before it was quenched
with a NH4Cl solution. After extraction with CH2Cl2 followed by
removal of the solvent under reduced pressure, the products were
purified by flash column chromatography (silica gel; n-hexane/
EtOAc, 90:10) to yield 15 (33%). 1H NMR (400 MHz, CDCl3,
3
25 °C): δ = 1.75 (s, 3 H), 3.65 (s, 3 H), 4.47 (d, JH,H = 14.8 Hz, 1
3
3
H), 4.67 (d, JH,H = 16.2 Hz, 1 H), 4.69 (d, JH,H = 14.8 Hz, 1 H),
3
4.88 (d, JH,H = 16.2 Hz, 1 H), 5.29 (s, 1 H), 7.10–7.12 (m, 2 H),
7.19–7.21 (m, 2 H), 7.28–7.39 (m, 6 H) ppm. 13C NMR (100 MHz,
CDCl3, 25 °C): δ = 28.0, 49.2, 50.8, 52.7, 66.5, 76.7, 85.7, 127.2,
127.8, 127.9, 128.0, 128.7, 128.8, 128.9, 134.6, 135.7, 170.7 ppm.
IR (CHCl ): ν = 3360.1, 3023.8, 2929.0, 2238.8, 1716.9, 1645.2,
˜
3
1436.3, 1363.3, 1261.4 cm–1. MS (70 eV): m/z (%): 333 (20) [M –
H2O]+, 224 (55), 127 (17), 95 (20), 92 (65), 91 (100), 65 (33).
HRMS: calcd. for C21H19NO3 [M – H2O]+ 333.1352; found
333.1365.
This mechanistic proposal diverges from that advanced
by Reboul et al.[1] in the fluoride-catalyzed synthesis of 1,3-
benzothiazines from alkyl propiolates and cyclic sulfon-
amides. The authors propose a catalytic cycle triggered and
maintained by fluoride ions. This proposal requires that the
fluoride ion must be regenerated during the catalytic cycle
to keep the process going. It is not easy to explain that if the 15.
fluoride ion can deprotonate the starting alkynoate, which
means that the fluoride is a stronger base than the gener-
ated acetylide, hydrogen fluoride can be deprotonated in the
presence of alkyl propiolate (hydrogen fluoride should be a
Supporting Information (see footnote on the first page of this arti-
cle): General experimental details and characterization data for
compounds 9–12 and 14; H and 13C NMR spectra of compound
1
Acknowledgments
This research was supported by the Spanish Ministerio de Ciencia
Innovación, the European Regional Development Fund
(CTQ2005-09074-C02-02 and CTQ2008-06806-C02-02), the Span-
ish MSC ISCIII (RETICS RD06/0020/1046), CSIC (Proyecto In-
tramural Especial 200719), FUNCIS (REDESFAC PI01/06 and 35/
milder acid than alkyl propiolate). We believe that a cata-
e
lytic mechanism involving fluoride triggering and autocata-
lytic maintenance would be more appropriate. Our own re-
sults confirm this idea.
36
www.eurjoc.org
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2010, 33–37

Fluoride-Triggered Domino Reactions
F. García-Tellado, Chem. Eur. J. 2005, 11, 3502–3510; d) D.
Tejedor, F. García-Tellado, J. J. Marrero-Tellado, P. de Armas,
Chem. Eur. J. 2003, 9, 3122–3131.
06), and the Fundación Instituto Canario de Investigación del
Cáncer (FICIC-G.I.N808/2007). S. L.-T. and G. M.-A. thank the
Ministerio de Educación y Ciencia (MEC) for their FPU grants.
The authors thank technician Ms. Anna Jurado Varona for her
experimental assistance.
[5] D. Tejedor, A. Santos-Expósito, F. García-Tellado, Chem.
Commun. 2006, 2667–2669.
[6] D. Tejedor, S. López-Tosco, J. González-Platas, F. García-Tel-
lado, J. Org. Chem. 2007, 72, 5454–5456.
[7] D. Tejedor, G. Méndez-Abt, J. González-Platas, M. A. Ramí-
rez, F. Garcia-Tellado, Chem. Commun. 2009, 2368–2370.
[8] For a study of catalyzed asymmetric aldol reactions of pyr-
uvates, see: N. Gathergood, K. Juhl, T. B. Poulse, K. Thordrup,
K. A. Jørgensen, Org. Biomol. Chem. 2004, 2, 1077–1085 and
references cited therein.
[9] I. Dickstein, S. I. Miller in The Chemistry of Functional Groups:
The Chemistry of The Carbon-Carbon Triple Bond, Part 2 (Ed.:
S. Patai), Wiley, Chichester, 1978, ch. 19, pp. 813–955.
[10] D. T. Aragón, G. V. López, F. Garcia-Tellado, J. J. Marrero-
Tellado, P. de Armas, D. Terrero, J. Org. Chem. 2003, 68, 3363–
3365.
[1]
[2]
C. Splitz, J.-F. Lohier, V. Reboul, P. Metzner, Org. Lett. 2009,
11, 2776–2779.
ABBЈ 3CR refers to a three-component reaction that utilizes
two different components (A and B) to give a product that
incorporates into its structure one unit of component A and
two chemodifferentiated units of component B (B and BЈ). For
full details and more examples of this type of multicomponent
reaction: D. Tejedor, F. García-Tellado, Chem. Soc. Rev. 2007,
36, 484–491.
[3] J. H. Clark, Chem. Rev. 1980, 80, 429–452.
[4] a) D. Tejedor, S. López-Tosco, F. Cruz-Acosta, G. Méndez-
Abt, F. García-Tellado, Angew. Chem. Int. Ed. 2009, 48, 2090–
2098; b) D. Tejedor, A. Santos-Expósito, F. García-Tellado,
Chem. Eur. J. 2007, 13, 1201–1209; c) D. Tejedor, D. González-
Cruz, A. Santos-Expósito, J. J. Marrero-Tellado, P. de Armas,
[11] K. Maruoka (Ed.), Asymmetric Phase Transfer Catalysis,
Wiley-VCH, Weinheim, 2008.
Received: September 22, 2009
Published Online: November 11, 2009
Eur. J. Org. Chem. 2010, 33–37
© 2010 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
37