Bioinspired metal-catalysed Doebner-Knoevenagel condensations between malonic acid half thioesters and aldehydes

Article abstract of DOI:10.1002/ejoc.200600880

Inspired by polyketide biosynthesis, unprecedented metal-catalysed Doebner-Knoevenagel condensations between Malonic Acid Half Thioesters (MAHTs) and various aldehydes have been achieved through the use of Yb(OTf)₃ as catalyst, mimicking a five

Full text of DOI:10.1002/ejoc.200600880

FULL PAPER
DOI: 10.1002/ejoc.200600880
Bioinspired Metal-Catalysed Doebner–Knoevenagel Condensations between
Malonic Acid Half Thioesters and Aldehydes
Fabrice Berrué,[a] Sylvain Antoniotti,[a] Olivier P. Thomas,*[a] and Philippe Amade[a]
Keywords: Polyketides / Biomimetic synthesis / Metal catalysis / Knoevenagel condensation
Inspired by polyketide biosynthesis, unprecedented metal-
catalysed Doebner–Knoevenagel condensations between
Malonic Acid Half Thioesters (MAHTs) and various alde-
hydes have been achieved through the use of Yb(OTf)3 as
catalyst, mimicking a five-domain polyketide synthase mod-
ule for the first time. In the presence of 5-methoxybenzimid-
azole, stereo- and regiocontrol of the reaction proved to be
high.
(© Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2007)
egy towards polyketide precursors based on a simple chemi-
cal iterative process satisfying the following requirements:
(1) mimicry of each PKS module of type I, II and III reac-
tions, (2) selective metal catalysis (M1, M2 and M3, respec-
tively), (3) mild conditions, and (4) no or few by-products.
Malonic Acid Half Thioesters (MAHTs) were selected as
carbon-centred nucleophiles because of their natural occur-
rence and their typical reactivity.
Introduction
Polyketides represent a vast class of natural products
with large chemical diversity and diverse biological activi-
ties.[1] The common features that link polyketides as a spe-
cific chemical class are the sequence of reactions they origi-
nate from, and the related intermediates of these reactions.
Each polyketide is produced by a unique PolyKetide Syn-
thase (PKS), made up of several enzymes or modules.[2] In
each module, all of which contain at least the three domains
Keto Synthase (KS), Acyl Transferase (AT), and Acyl Car-
rier Protein (ACP), the reaction sequence begins with a de-
carboxylative Claisen condensation between two thioesters
(derived from acetyl-CoA) to afford the corresponding β-
ketothioester. In more complex modules the condensation
step can be followed by a reduction process induced by the
insertion of a Keto Reductase (KR) domain to yield a β-
hydroxy thioester. If the reduction process includes both
KR and an additional DeHydratase (DH) domain, subse-
quent dehydration of the β-hydroxy thioester affords an un-
saturated thioester in either the α,β or the β,γ positions
(Scheme 1).[3]
Using such intermediates, Kobuke and co-workers dem-
onstrated thirty years ago that a decarboxylative Claisen
condensation, corresponding to the type I reaction, was
feasible through the use of MgII catalysis (M1) and imid-
azole in an enzyme-free system.[6] The Kobuke reactions
contrast with the harsher conditions required for other syn-
thetic approaches to polyketide synthesis. They were re-
cently improved upon by Matile and co-workers, who dis-
covered that benzimidazole derivatives in association with
MgII salts can serve as efficient base catalysts in this reac-
tion.[7] Very recently, Shair and co-workers described a mild
and catalytic thioester aldol reaction of type II, involving
CuII catalysis (M2) and aldehydes as substrates,[8] whilst the
same aldolisation was also shown to proceed with simple
thioesters and stoichiometric MgII.[9] Inspired by these ef-
ficient synthetic strategies, we focused on the type III reac-
tion, known as the Doebner–Knoevenagel condensation,
and largely applied to malonic acid half oxoesters.[10] Mild
conditions using 4-(dimethylamino)pyridine (DMAP) catal-
ysis have been described only recently,[11] but no simple pro-
cedure for MAHTs has been proposed to date. Here we
report the first metal-catalysed Doebner–Knoevenagel con-
densations between MAHTs and aldehydes, mimicking a
five-domain module (Scheme 1). This reaction offers a
route to iterative metal-catalysed condensations of thioes-
ters to afford complex polyketides.
Inspired by this biosynthetic pathway, a variety of car-
bon–carbon bond-forming reactions that enable the synthe-
sis of complex polyketides have been developed.[4] Never-
theless, these successive reactions are often time- and rea-
gent-consuming and always require the use of reactive inter-
mediates that yield various by-products.[5] We therefore de-
cided to devote our efforts to a bioinspired chemical strat-
[a] Laboratoire de Chimie des Molécules Bioactives et des Arômes
UMR 6001 CNRS, Institut de Chimie de Nice, Faculté des
Sciences, Université de Nice-Sophia Antipolis,
Parc Valrose, 06108 Nice Cedex 2, France
Fax: +33-4-92076599
E-mail: olivier.thomas@unice.fr
Eur. J. Org. Chem. 2007, 1743–1748
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1743

F. Berrué, S. Antoniotti, O. P. Thomas, P. Amade
FULL PAPER
Scheme 1. Metal-catalysed reactions mimicking PKS modules.
Table 1. Influence of the base on the condensation between MAHT
2 and phenylacetaldehyde (1).
Results and Discussion
Phenylacetaldehyde (1) was chosen as a substrate in our
preliminary investigations because of its natural occurrence
and its putative role as a precursor of various polyketides,
including some new spiculoic acids that we had isolated and
characterized from the marine sponge Plakortis zyggom-
pha.[12] Our interest came from the postulated biosynthetic
pathway of these metabolites, in which the dehydration step
is believed to occur unexpectedly in the β,γ-position of the
additional unit. The first attempts with phenylacetaldehyde
(1) involved combining CuII 2-ethylhexanoate (20 mol-%),
5-methoxybenzimidazole (22 mol-%) and S-benzyl MAHT
(2) in THF at room temperature.[8] In good agreement with
Shair’s results, only the β-hydroxy thioester 3 was formed
under these conditions. After a screening of various metallic
salts, the first catalytic Doebner–Knoevenagel condensation
of MAHT was achieved with 10 mol-% of Zn(OTf)2 or
Mg(OTf)2 to afford, in moderate 45% and 55% yields, mix-
tures of the α,β- and β,γ-unsaturated thioesters 4 and 5 in
1:1 and 3:7 ratios, respectively. Remarkably, no formation
of the β-hydroxy thioester 3 was observed. A more interest-
ing result was obtained with the use of a catalytic amount
of Yb(OTf)3 in the presence of a stoichiometric amount of
5-methoxybenzimidazole. Indeed, the yield was raised to
75% with this catalyst, whilst the regioselectivity was higher
than 95% in favour of the biomimetic β,γ-unsaturated
product 5. This satisfactory outcome with Yb(OTf)3
prompted us to study the influence of other parameters, in
particular the role of the base (Table 1).
Entry
Base
% isol. yield
4/5[a]
1
2
3
4
5
6
7
8
5-methoxybenzimidazole
imidazole
triethylamine
2,6-lutidine
N-methylmorpholine
DBU
2,6-di-tert-butylpyridine
4-(dimethylamino)pyridine
5-methoxybenzimidazole
2,6-lutidine
75
37
18
98
63
–
35
–
10
35
Ͻ5:Ͼ95
22:78
11:89
Ͻ5:Ͼ95
11:89
–
Ͻ5:Ͼ95
–
41:59
Ͻ5:Ͼ95
9[b]
10[b]
1
[a] Determined by H NMR spectroscopy. [b] No catalyst, in 68 h.
Every attempt to reduce the amount of base always re- interesting result was obtained with 2,6-lutidine, which se-
sulted in a significant decrease in the reaction yield. These lectively afforded the biomimetic β,γ-unsaturated thioester
results underlined the subtle conditions necessary for this 5 in quantitative yield. In order to assess the role of the
condensation to be efficient. Indeed, poor results were ob- base in the reaction mechanism, experiments in the absence
tained with the stronger bases Et3N, DBU and DMAP of metal catalyst were conducted with 5-methoxybenzimid-
(Table 1, Entries 3, 6, and 8), the only product detectable azole and 2,6-lutidine, which had previously given the best
under these conditions being the S-benzyl thioester acetate, results. The yield of the reaction was drastically decreased,
a product of MAHT 2 decarboxylation, isolated in quanti- even after 68 h reaction time, which underlines the catalytic
tative yield in the presence of DMAP. This result was unex- role of the metal. Moreover, with 5-methoxybenzimidazole
pected, since DMAP is known to catalyse direct aldol ad- (Table 1, Entry 9) the regioselectivity almost disappeared,
ditions of malonic acid half oxoesters.[11] On comparison of but with 2,6-lutidine (Table 1, Entry 10) compound 5 re-
the results obtained with 2,6-lutidine and 2,6-di-tert-butyl- mained the only product of the reaction. It could be specu-
pyridine (Table 1, Entries 4 and 7), it is clear that steric ef- lated that the catalysed reaction could involve chelation of
fects are prone to alter the course of the reaction. A very the metal with the leaving hydroxy group and the thioester
1744
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 1743–1748

Bioinspired Metal-Catalysed Doebner–Knoevenagel Condensations
FULL PAPER
carbonyl in a putative six-membered ring transition state. Table 2. Catalytic Doebner–Knoevenagel condensations between
MAHT 2 and various aldehydes.[a]
Such activation should induce the higher yields and shorter
reaction times observed in the catalysed reaction.
In order to evaluate the relative thermodynamic stabili-
ties of the two isomers 4 and 5, the dehydration step was
performed under standard synthetic conditions, starting
from the β-hydroxy thioester 3. Dehydration was achieved
through the formation of the methanesulfonyl ester by
treatment with MsCl in CH2Cl2/pyridine (9:1) to afford the
unsaturated compounds 4 and 5 in a 9:1 ratio. When the
reaction was carried out in [D5]pyridine at 20 °C, however,
1
the composition of the mixture as monitored by H NMR
evolved, ending up after 6 h with pure compound 5 as the
thermodynamic product of the reaction. This result demon-
strated that pyridine and derivatives are basic enough to
isomerise 4 into the thermodynamically more stable com-
pound 5 whereas 5-methoxybenzimidazole is less efficient
in this regard. Two distinct mechanisms leading to 5 were
thus demonstrated for the catalytic Doebner–Knoevenagel
condensation: (1) addition/dehydration to give 4, followed
by isomerisation into 5 in the presence of pyridine deriva-
tives, and (2) addition/dehydration to afford 5 directly in the
presence of 5-methoxybenzimidazole.
Using both sets of conditions (2,6-lutidine and 5-meth-
oxybenzimidazole), we investigated the scope of the reac-
tion with a variety of aldehydes as potential polyketide bio-
synthesis starter units. Surprisingly, the production of the
dehydrated product was much more efficient with 5-meth-
oxybenzimidazole in all cases; with saturated aliphatic alde-
[a] Conducted under N2 by combining 1:1:1 equiv. of aldehyde,
hydes the dehydrated products were not even isolated when
2,6-lutidine was used. We then focused on 5-methoxybenz-
imidazole as a base for further experiments (Table 2).
With arylacetaldehydes in the presence of 5-methoxy-
MAHT
2 and 5-methoxybenzimidazole in the presence of
Yb(OTf)3 (10 mol-%) in THF at room temperature in 16 h.
benzimidazole, conversion into unsaturated thioesters oc- salts, the first putative β-hydroxy intermediate A undergoes
curred in high yields (Table 2, Entries 1–2), with the regiose- a fast decarboxylative process to afford the β-hydroxy thio-
lectivity always in favour of the β,γ-conjugated thioesters. ester 3, as described by Shair and co-workers (Scheme 2).[8]
Interestingly, the reaction was also amenable to the use of
With zinc, magnesium and ytterbium salts, it is assumed
an aliphatic β,γ-unsaturated aldehyde with the same effi- that the dehydration is faster than the decarboxylation pro-
ciency and again with the same regioselectivity, involving cess, affording the α,β-conjugated intermediate B1. Iso-
conjugation with the cyclohexenyl substituent (Table 2, En- merisation followed by decarboxylation would then give
try 3). The case of 2-phenylpropionaldehyde appeared to be product 5. The fact that 3 is not an intermediate in the
an exception, with a yield reduced to 35% and a reverse formation of 4 and 5 was demonstrated unequivocally, as
regioselectivity in favour of the α,β-unsaturated product no reaction occurred when 3 was subjected to our opti-
(Table 2, Entry 4), though the same trend was also observed mized conditions [Yb(OTf)3, 5-methoxybenzimidazole in
with saturated aliphatic aldehydes and arylcarboxaldehydes THF at room temp.]. The transient existence of the reactive
(Table 2, Entries 5–10). The strongly electron-withdrawing intermediate B1 was supported by the fact that no dehy-
effect of para nitro substitution on the aromatic ring even drated product was observed with α-methyl-substituted
resulted in total inhibition of the reaction, while the meta MAHT 2Ј (Scheme 3).
isomer reacted to give a 31% yield (Table 2, Entries 9–10).
In this case the reaction stopped after the decarbox-
It is worth noting that the stereoselectivity was always to- ylative Claisen addition, providing the β-hydroxy thioester
tally in favour of the (E) stereoisomer, as shown by the large 3Ј in 65% yield, just as in a type II reaction. The quaternary
1
coupling constants of the vinylic protons in the H NMR carbon atom at the α-position of the thioester intermediate
spectra of the reaction products.
AЈ would indeed make a dehydration process to afford the
Subsequent experiments were performed in order to gain α,β-unsaturated intermediate B1Ј impossible. Furthermore,
new insights into the reaction mechanism. When all rea- a decarboxylative isomerisation process involving B1, as
gents with the exception of the aldehyde were mixed to- previously described by Corey in the case of unsaturated
gether, no decarboxylation occurred, suggesting that the ad- malonic acid derivatives,[13] would by itself explain the pres-
dition is not induced by prior decarboxylation. With copper ence of the mixtures of compounds 4 and 5, in ratios dic-
Eur. J. Org. Chem. 2007, 1743–1748
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1745

F. Berrué, S. Antoniotti, O. P. Thomas, P. Amade
FULL PAPER
Scheme 2. Mechanistic insights into the Doebner–Knoevenagel condensations between MAHT and aldehydes.
starting point in polyketide synthesis. Indeed, an iterative
process using the easy reduction of thioesters into aldehydes
and additional condensations using the same as electro-
philes can be considered for future extensions.[14]
Scheme 3. Reaction between the α-methyl-substituted MAHT 2Ј
and phenylacetaldehyde (1).
Experimental Section
General: All chemicals were obtained from commercial suppliers
and were used without further purification. All reactions were car-
ried out under nitrogen with use of standard Schlenk techniques.
THF was distilled from sodium/benzophenone. 1H and 13C spectra
were recorded on either a Bruker Avance 500 MHz or an Avance
tated by the natures of the metal catalysis and the sub-
strates. The presence of the phenyl-conjugated intermediate
B2 in the mechanistic pathway leading to 5 was, however,
demonstrated by an additional experiment. With an excess
of aldehyde 1 rather than a stoichiometric quantity, we ob-
served the formation of a large amount of the adduct 6, the
result of the addition of B2 to a second molecule of alde-
hyde 1 (Scheme 2). This double addition was not observed
when product 5 was subjected to the same reaction condi-
tions. On the basis of the mechanistic proposal, we assume
that the outcome of the reaction (yield and regioselectivity)
is governed by the relative thermodynamic stabilities of the
B1- and B2-derived carbanions. With a poorly stabilized B1
carbanion the yield drops because of a slow decarboxylative
process on B1 (Table 2, Entries 4–8), whilst with a strongly
stabilized B1 carbanion the regioselectivity in favour of the
β,γ-unsaturated thioester decreases (Table 2, Entry 2).
200 MHz NMR spectrometer with CDCl3 (δH = 7.26 and δC
=
77.16 ppm) as internal standards. Spectral features are tabulated in
the following order: chemical shift (δ in ppm), multiplicity (s =
singlet, d = doublet, t = triplet, q = quartet, h = heptet, m = com-
plex multiplet), number of protons, coupling constant (J in Hz),
assignment. ElectroSpray Ionisation (ESI) mass spectra were run
on a Bruker Esquire 3000 Plus spectrometer in the positive or nega-
tive mode. High-resolution mass spectra were obtained with a LCT
Waters-Micromass mass spectrometer (ESI TOF). IR spectra were
obtained with a Perkin–Elmer Paragon 1000 FT-IR spectrometer.
General Procedure for the Preparation of Malonic Acid Half Thioes-
ters: The malonic or methylmalonic acid (50 mmol, 4 equiv.) was
added to a stirred solution of the polyphosphate ester (PPE, 7.8 g)
in CHCl3 (65 mL) and THF (15 mL), and benzyl mercaptan
(1.46 mL, 12.5 mmol, 1 equiv.) diluted in CHCl3 (20 mL) was then
added to the resulting white suspension by syringe. The reaction
mixture was stirred at 25 °C for 16 h and the quenched with aque-
ous HCl (0.5 , 200 mL). The mixture was extracted with CH2Cl2
(2ϫ200 mL) and the organic layer was dried with MgSO4, filtered
and concentrated under reduced pressure. Purification of the resi-
due by flash chromatography on silica gel [n-hex/EtOAc (8:2), then
n-hex/EtOAc (5:5)] afforded the pure malonic acid half thioesters.
Conclusions
We have developed original, bioinspired, metal-catalysed
Doebner–Knoevenagel condensations between MAHTs
and aldehydes. The reactions can be conducted under very
mild conditions with catalytic quantities of Yb(OTf)3 in the
presence of stoichiometric amounts of 5-methoxybenzimid-
azole and always proceed with high regio- and stereoselecti-
vity. The simple reaction conditions and the accessibility of
S-Benzyl Malonic Acid Half Thioester 2: White solid; purified by
flash chromatography; 1.77 g or 68% isolated yield. 1H NMR
(500 MHz, CDCl3): δ = 9.75 (brs, OH), 7.14–7.40 (m, 5 H, Ph),
MAHTs are attractive incentives to use this reaction as a 4.16 (s, 2 H, H-4), 3.61 (s, 2 H, H-2) ppm. 13C NMR (125 MHz,
1746
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 1743–1748

Bioinspired Metal-Catalysed Doebner–Knoevenagel Condensations
CDCl3): δ = 190.9 (C-3), 170.9 (C-1), 136.5, 129.0, 128.9, 127.7,
FULL PAPER
16.0 Hz, 1 H, 3-H), 7.30–7.07 (m, Ph), 6.01 (dt, JH,H = 15.5,
1.5 Hz, 1 H, 2-H), 4.09 (s, 2 H, 1Ј-H), 3.75 (dd, 3JH,H = 6.5, 1.5 Hz,
2 H, 4-H) ppm. 13C NMR (125 MHz, CDCl3): δ = 189.0 (C-1),
149.0, 141.5 (C-3), 137.6, 133.6, 132.4 (C-2), 129.9, 128.9, 128.7,
128.3, 127.4, 125.3, 35.7 (C-4), 33.2 (C-1Ј) ppm. MS (ESI, MeOH):
m/z: 336 [M + Na]+.
3
48.6 (C-2), 34.1 (C-4) ppm. IR: ν = 3453 (OH), 1726 (CO), 1687
˜
(CO), 1496, 1454 cm–1. MS (ESI, MeOH): m/z: 209 [M – H]–.
S-Benzyl Methylmalonic Acid Half Thioester 2Ј: Colourless oil;
1
purified by flash chromatography; 1.79 g or 64% isolated yield. H
NMR (500 MHz, CDCl3): δ = 11.05 (brs, OH), 7.25–7.40 (m, 5 H,
Ph), 4.23 (m, 2 H, H-4), 3.75 (q, 1 H, J = 7.3 Hz, H-2), 1.52 (d, 3
H, J = 7.3 Hz, H-5) ppm. 13C NMR (125 MHz, CDCl3): δ = 195.1
(C-3), 174.8 (C-1), 136.6, 128.8, 128.7, 127.5, 53.5 (C-2), 33.7 (C-
4), 14.1 (C-5) ppm. MS (ESI, MeOH): m/z: 223 [M – H]–.
β,γ-Unsaturated Thioester: 1H NMR (500 MHz, CDCl3): δ = 7.86
3
(brd, JH,H = 8.0 Hz, 1 H), 7.60–7.40 (m, Ph), 7.30–7.07 (m, Ph),
3
3
6.96 (d, JH,H = 15.0 Hz, 1 H, 4-H), 6.21 (dt, JH,H = 15, 7.5 Hz,
3
1 H, 3-H), 4.08 (s, 2 H, 1Ј-H), 3.47 (dd, JH,H = 7.5, 1.5 Hz, 2 H,
2-H) ppm. 13C NMR (125 MHz, CDCl3): δ = 196.3 (C-1), 147.7,
137.3, 133.3, 130.2 (C-4), 129.0, 128.8, 128.4, 127.5, 126.5, 124.7
(C-3), 47.5 (C-2), 33.6 (C-1Ј) ppm. MS (ESI, MeOH): m/z: 336 [M
+ Na]+.
Procedure for the Catalytic Aldol Reactions between MAHTs and
Phenylacetaldehyde (1): 5-Methoxybenzimidazole (0.057 mmol,
0.22 equiv.) and Cu(2-ethylhexanoate)2 (0.052 mmol, 0.20 equiv.)
were added at 25 °C to a stirred solution of S-benzyl malonic acid
half thioester 2 (55 mg, 0.26 mmol) in THF (5 mL). Phenylacetal-
dehyde (1, 0.26 mmol, 1 equiv.) was added after homogenisation of
the solution (≈ 2 min), which was then stirred for 16 h at 25 °C and
quenched with an aqueous solution of HCl (0.5 , 10 mL). The
resulting mixture was diluted with EtOAc and washed successively
with aq. HCl (0.5 ), saturated aq. NaHCO3 and brine. The or-
ganic layer was dried with MgSO4, filtered and concentrated under
reduced pressure. The product was purified by flash column
chromatography on silica gel (from n-hex to EtOAc) to afford pure
compound 3.
S-Benzyl
(E)-4-(2,6,6-Trimethylcyclohex-1-enyl)but-3-enethioate
(Table 2, Entry 3): Colourless oil; purified by flash chromatog-
1
raphy; 67 mg or 82% isolated yield. H NMR (200 MHz, CDCl3):
3
δ = 7.09–7.28 (m, 5 H, Ph), 5.98 (d, JH,H = 15.4 Hz, 1 H, 4-H),
3
5.40 (dt, JH,H = 15.4 7.2 Hz, 1 H, 3-H), 4.21 (s, 2 H, 1Ј-H), 3.28
(d, 3JH,H = 7.2 Hz, 2 H, 2-H), 1.90 (m, 2 H, H-3ЈЈ), 1.65–1.28 (m, 7
H, 4ЈЈ,5ЈЈ,9ЈЈ-H), 0.91 (s, 6 H, 7ЈЈ,8ЈЈ-H) ppm. 13C NMR (50 MHz,
CDCl3): δ = 197.6 (C-1), 137.9, 137.0 (C-1ЈЈ), 134.0 (C-4), 129.5
(C-2ЈЈ), 128.9, 128.7, 127.3, 124.5 (C-3), 48.1 (C-2), 39.5 (C-5ЈЈ),
34.2 (C-6ЈЈ), 33.4 (C-1Ј), 32.8 (C-3ЈЈ), 28.8 (C-7ЈЈ, C-8ЈЈ), 21.6 (C-
4ЈЈ), 19.4 (C-9ЈЈ) ppm. MS (ESI, MeOH): m/z: 337 [M + Na]+.
HRMS (ESI, MeOH): m/z: 337.1616 [M + Na]+; 337.1602 calcd.
for C20H26ONaS.
S-Benzyl 3-Hydroxy-4-Phenylbutanethioate (3): Colourless oil;
48 mg or 48% isolated yield. 1H NMR (500 MHz, CDCl3): δ =
3
7.15–7.34 (m, 10 H, Ph), 4.25 (m, 1 H, 3-H), 4.08 (d, JH,H
=
3
13.8 Hz, 1 H, 1Јa-H), 4.04 (d, JH,H = 14.1 Hz, 1 H, 1Јb-H), 2.70
(m, 2 H, 4-H), 2.66 (m, 2 H, 2-H) ppm. 13C NMR (125 MHz,
CDCl3): δ = 198.2 (C-1), 137.4, 137.1, 129.4, 128.7, 128.6, 128.5,
127.3, 126.7, 69.6 (C-3), 49.6 (C-2), 42.9 (C-4), 33.3 (C-1Ј) ppm.
S-Benzyl (E)-4-Phenylpent-2-enethioate (Table 2, Entry 4): Colour-
less oil; purified by flash chromatography; 25 mg or 35% isolated
1
yield. H NMR (500 MHz, CDCl3): δ = 7.09–7.25 (m, 10 H, Ph),
3
3
7.00 (dd, JH,H = 15.6, 6.4 Hz, 1 H, 3-H), 6.01 (dd, JH,H = 15.6,
1.8 Hz, 1 H, 2-H), 4.10 (s, 2 H, 1Ј-H), 3.54 (m, 1 H, 4-H), 1.36 (d,
3JH,H = 6.4 Hz, 3 H, 5-H) ppm. 13C NMR (125 MHz, CDCl3): δ =
189.5 (C-1), 149.3 (C-3), 143.1, 137.8, 129.0, 128.9, 128.7, 127.5,
IR: ν = 3540 (OH), 1686 (CO), 1495, 1453 cm–1. MS (ESI, pos,
˜
MeOH): m/z: 309 [M + Na]+. HRMS (ESI, pos, MeOH): m/z:
309.0928 [M + Na]+; 309.0925 calcd. for C17H18O2NaS.
General Procedure for the Yb(OTf)3-Catalyzed Doebner–Knoevena-
gel Condensation between MAHTs and Aldehydes: The base
(0.26 mmol, 1 equiv.) and Yb(OTf)3 (0.026 mmol, 10 mol-%) were
added at 25 °C to a stirred solution of S-benzyl malonic acid half
thioester 2 (55 mg, 0.26 mmol) in THF (5 mL). The aldehyde
(0.26 mmol, 1 equiv.) was added after homogenization of the solu-
tion (≈ 2 min), which was then stirred for 16 h at 25 °C and concen-
trated under reduced pressure to yield a yellow oil. The oil was
purified by flash chromatography on silica gel [n-hexane/CH2Cl2
(1:1) then CH2Cl2 (100%)] to afford the following unsaturated thio-
esters.
127.4, 127.0 (C-2), 42.2 (C-4), 21.9 (C-5), 37.4 (C-1Ј) ppm. IR: ν =
˜
1671 (CO), 1627, 1493, 1452 cm–1. MS (ESI, MeOH): m/z: 305 [M
+ Na]+. HRMS (ESI, MeOH): m/z: 305.1014 [M + Na]+; 305.0996
calcd. for C18H18OSNa.
S-Benzyl (E)-But-2-enethioate (Table 2, Entry 5): Colourless oil;
purified by flash chromatography; 17 mg or 34% isolated yield. 1H
NMR (500 MHz, CDCl3): δ = 7.16–7.24 (m, 5 H, Ph), 6.86 (dq,
3
3JH,H = 15.6, 6.9 Hz, 1 H, 3-H), 6.07 (dq, JH,H = 15.6, 1.8 Hz, 1
3
H, 2-H), 4.11 (s, 2 H, 1Ј-H), 1.80 (dd, JH,H = 6.9, 1.8 Hz, 3 H, 4-
H) ppm. 13C NMR (125 MHz, CDCl3): δ = 189.3 (C-1), 141.4 (C-
3), 137.9, 129.9 (C-2), 129.0, 128.7, 127.3, 33.0 (C-1Ј), 18.1 (C-
S-Benzyl (E)-4-Phenylbut-3-enethioate (5, Table 2, Entry 1):
4) ppm. IR: ν = 1671 (CO), 1630 cm–1. MS (ESI, MeOH): m/z: 215
˜
Colourless oil; purified by flash chromatography; 52 mg or 75%
[M + Na]+. HRMS (ESI, MeOH): m/z: 215.0489 [M + Na]+;
1
isolated yield. H NMR (500 MHz, CDCl3): δ = 7.15–7.50 (m, 10
215.0507 calcd. for C11H12ONaS.
H, Ph), 6.65 (d, 3JH,H = 16.0 Hz, 1 H, 4-H), 6.40 (dt, 3JH,H = 16.0,
3
7.3 Hz, 1 H, 3-H), 4.26 (s, 2 H, 1Ј-H), 3.57 (d, JH,H = 7.3 Hz, 2
S-Benzyl (E)-Pent-2-enethioate (Table 2, Entry 6): Colourless oil;
purified by flash chromatography; 19 mg or 35% isolated yield. 1H
NMR (500 MHz, CDCl3): δ = 7.16–7.24 (m, 5 H, Ph), 6.90 (dt,
H, H-2) ppm. 13C NMR (125 MHz, CDCl3): δ = 196.9 (C-1), 137.5,
136.7, 135.1 (C-4), 129.0, 128.7, 127.9, 127.4, 126.5, 120.9 (C-3),
3
˜
3JH,H = 15.6, 6.4 Hz, 1 H, 3-H), 6.04 (dt, JH,H = 15.6, 1.8 Hz, 1
47.5 (C-2), 33.5 (C-1Ј) ppm. IR: ν = 1683 (CO), 1648, 1494,
1400 cm–1. MS (ESI, MeOH): m/z: 269 [M + H]+. HRMS (ESI,
3
H, 2-H), 4.12 (s, 2 H, 1Ј-H), 2.16 (qdd, JH,H = 6.9, 6.4, 1.8 Hz, 2
+
Na]+; 291.0820 calcd. for
3
H, 4-H), 0.99 (t, JH,H = 6.8 Hz, 3 H, 5-H) ppm. 13C NMR
MeOH): m/z: 291.0814 [M
C17H16ONaS.
(125 MHz, CDCl3): δ = 189.5 (C-1), 147.5 (C-3), 137.9, 129.0,
128.7, 127.5 (C-2), 127.3, 33.0 (C-1Ј), 25.5 (C-4), 12.2 (C-5) ppm.
S-Benzyl (E)-4-(2-Nitrophenyl)but-3-enethioate and S-Benzyl (E)-4-
(2-Nitrophenyl)but-2-enethioate (Table 2, Entry 2): Colourless oil;
purified by flash chromatography; 74 mg or 91% isolated yield in
a 7:3 mixture.
IR: ν = 1673 (CO), 1631 cm–1. MS (ESI, MeOH): m/z: 229 [M +
˜
Na]+. HRMS (ESI, MeOH): m/z: 229.0682 [M + Na]+; 229.0663
calcd. for C12H14ONaS.
1
α,β-Unsaturated Thioester: H NMR (500 MHz, CDCl3): δ = 7.95
S-Benzyl (E)-Non-2-enethioate (Table 2, Entry 7): Colourless oil;
(dd, 3JH,H = 8.0, 1.5 Hz, 1 H), 7.60–7.40 (m, Ph), 7.36 (m, JH,H
=
purified by flash chromatography; 22 mg or 32% isolated yield. 1H
3
Eur. J. Org. Chem. 2007, 1743–1748
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
1747

F. Berrué, S. Antoniotti, O. P. Thomas, P. Amade
FULL PAPER
NMR (500 MHz, CDCl3): δ = 7.30–7.10 (m, 5 H, Ph), 6.86 (dt, dried with MgSO4, filtered and concentrated under reduced pres-
3
3JH,H = 15.5, 6.5 Hz, 1 H, 3-H), 6.04 (dt, JH,H = 15.5, 1.5 Hz, 1 sure to yield 45 mg of a mixture of compounds 4 and 5 along with
3
H, 2-H), 4.11 (s, 2 H, 1Ј-H), 1.60–1.10 (m, 10 H), 0.81 (t, JH,H
=
the non-eliminated methanesulfonyl derivative. When the mixture
of compounds was dissolved in [D5]pyridine the elimination/isom-
7.0 Hz, 3 H, 9-H) ppm. 13C NMR (125 MHz, CDCl3): δ = 189.5
(C-1), 147.5 (C-3), 137.9, 129.0, 128.7, 127.5 (C-2), 127.3, 33.0 (C- erization process was followed by 1H NMR spectroscopy. After 6 h
1Ј), 25.5 (C-4), 12.2 (C-5) ppm. MS (ESI, MeOH): m/z: 285 [M +
no trace of compound 4 was present in the mixture.
Na]+.
S-Benzyl (E)-4-Methylpent-2-enethioate (Table 2, Entry 8): Colour- Acknowledgments
less oil; purified by flash chromatography; 22 mg or 38% isolated
yield. 1H NMR (500 MHz, CDCl3): δ = 7.15–7.25 (m, 5 H, Ph),
6.82 (dd, 3JH,H = 15.6 Hz, 1 H, 3-H), 6.00 (dd, 3JH,H = 15.6, 1.4 Hz,
1 H, 2-H), 4.11 (s, 2 H, 1Ј-H), 2.37 (hd, 3JH,H = 6.9, 1.4 Hz, 1 H, 4-
H), 0.99 (d, 3JH,H = 6.9 Hz, 6 H, 5,6-H) ppm. 13C NMR (125 MHz,
CDCl3): δ = 189.8 (C-1), 152.2 (C-3), 137.9, 129.0, 128.7, 127.3,
FB is grateful to the Region Provence-Alpes-Côte d’Azur for a fel-
lowship.
[1] J. Rohr, Angew. Chem. Int. Ed. 2000, 39, 2847–2849.
[2] J. Staunton, K. J. Weissman, Nat. Prod. Rep. 2001, 18, 380–
416.
[3] R. J. Heath, C. O. Rock, Nat. Prod. Rep. 2002, 19, 581–596.
[4] C. Le Sann, D. M. Muñoz, N. Saunders, T. J. Simpson, D. I.
Smith, F. Soulas, P. Watts, C. L. Willis, Org. Biomol. Chem.
2005, 3, 1719–1728.
125.7 (C-2), 33.3 (C-1Ј), 31.1 (C-4), 21.3 (C-5, C-6) ppm. IR: ν =
˜
1673 (CO), 1629, 1495, 1453 cm–1. MS (ESI, MeOH): m/z: 243 [M
+ Na]+. HRMS (ESI, MeOH): m/z: 243.0833 [M + Na]+; 243.0820
calcd. for C13H16ONaS.
[5] I. Paterson, C. J. Cowden, D. J. Wallace, Stereoselective Aldol
Reactions, in: the Synthesis of Polyketide Natural Products; in
Modern Carbonyl Chemistry (Ed.: J. Otera), Wiley-VCH,
Weinheim, 2000, pp. 249–298.
[6] Y. Kobuke, J. Yoshida, Tetrahedron Lett. 1978, 19, 367–370.
[7] N. Sakai, N. Sordé, S. Matile, Molecules 2001, 6, 845–851.
[8] D. Magdziak, G. Lalic, H. M. Lee, K. C. Fortner, A. D. Aloise,
M. D. Shair, J. Am. Chem. Soc. 2005, 127, 7284–7285, and ref-
erences therein.
S-Benzyl (E)-3-(3-Nitrophenyl)prop-2-enethioate (Table 2, Entry 9):
Colourless oil; purified by flash chromatography; 24 mg or 31%
isolated yield. 1H NMR (500 MHz, CDCl3): δ = 8.31 (s, 2ЈЈ-H),
8.16 (d, JH,H = 8.7 Hz, 1 H, 4ЈЈ-H), 7.74 (d, JH,H = 8.7 Hz, 1 H,
6ЈЈ-H), 7.58 (d, 3JH,H = 15.6 Hz, 1 H, 3-H), 7.50 (t, 3JH,H = 8.7 Hz,
1 H, 5ЈЈ-H), 7.17–7.28 (m, 5 H, Ph), 6.73 (d, 3JH,H = 15.6 Hz, 1 H,
2-H), 4.21 (s, 2 H, 1Ј-H) ppm. 13C NMR (125 MHz, CDCl3): δ =
188.7 (C-1), 148.8 (C-3ЈЈ), 137.8 (C-3), 137.3, 136.0 (C-1ЈЈ), 134.1
(C-6ЈЈ), 130.2 (C-5ЈЈ), 129.0, 127.8, 127.6 (C-2ЈЈ), 127.4, 124.9 (C-
3
3
[9] J. M. Yost, G. Zhou, D. M. Coltart, Org. Lett. 2006, 8, 1503–
1506.
4ЈЈ), 122.7 (C-2), 33.6 (C-1Ј) ppm. IR: ν = 1670 (CO), 1619, 1530
˜
(NO), 1350 (NO) cm–1. MS (ESI, MeOH): m/z: 317 [M + NH4]+.
HRMS (ESI, MeOH): m/z: 322.0534 [M + Na]+; 322.0514 calcd.
for C16H13NO3NaS.
[10] a) N. Ragoussis, V. Ragoussis, J. Chem. Soc., Perkin Trans. 1
1998, 3529–3540; b) Y. Peng, G. Song, Green Chem. 2003, 5,
704–707.
[11] B. List, A. Doehring, M. T. Hechavarria, K. Wobser, H.
van Thienen, R. Rios Torres, P. Llamas Galilea, Adv. Synth.
Catal. 2005, 347, 1558–1560.
[12] F. Berrué, O. P. Thomas, R. Fernandez, P. Amade, J. Nat. Prod.
2005, 68, 547–549.
[13] E. J. Corey, J. Am. Chem. Soc. 1953, 75, 1163–1167.
[14] T. Fukuyama, H. Tokuyama, Aldrichimica Acta 2004, 37, 87–
96.
Dehydration of 3 via a Methanesulfonyl Ester: Methanesulfonyl
chloride (0.85 mmol, 5 equiv.) was added at 25 °C to a stirred solu-
tion of the β-hydroxy thioester 3 (48 mg, 0.17 mmol) in a mixture
of pyridine (1 mL) and CH2Cl2 (9 mL). The solution was then
stirred for 16 h at room temperature and quenched with saturated
aq. NaHCO3. After extraction with CH2Cl2 (2ϫ20 mL), the or-
ganic phase was washed successively with aq. HCl (0.5  , 20 mL),
saturated aq. NaHCO3 (20 mL) and brine (20 mL). It was then
Received: October 8, 2006
Published Online: February 14, 2007
1748
www.eurjoc.org
© 2007 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 2007, 1743–1748