A versatile and highly stereoselective access to vinyl triflates derived from 1,3-dicarbonyl and related compounds

Article abstract of DOI:10.1021/jo8015049

(Chemical Equation Presented) 1,3-Dicarbonyl derivatives, such as 1,3-diketones, β-ketoaldehydes, β-ketoesters, β-ketoamides, β-ketophosphonates and β-ketosulfones were efficiently converted to the corresponding Z vinyl triflates with high stereoselectivi

Full text of DOI:10.1021/jo8015049

SCHEME 1. Two Conventional Methods for the Synthesis
of Vinyl Triflates from Ketones
A Versatile and Highly Stereoselective Access to
Vinyl Triflates Derived from 1,3-Dicarbonyl and
Related Compounds
Simon Specklin, Philippe Bertus,† Jean-Marc Weibel, and
Patrick Pale*
Laboratoire de Synthe`se et Re´actiVite´ Organiques, associe´
au CNRS, Institut de Chimie, UniVersite´ Louis Pasteur, 4 rue
Blaise Pascal, 67000 Strasbourg, France
ppale@chimie.u-strasbg.fr
ReceiVed July 9, 2008
SCHEME 2. Direct Route to Vinyl Triflates from Alkynes
ing a different mechanistic pathway (Scheme 1). One of them
is based on enolization with strong bases followed by trapping
with triflic anhydride or equivalents.1,5 In this sequence, the
enolization step controls the regio- and stereoselectivity of the
overall process. On the other hand, triflic anhydride addition to
ketones provides gem-bis-triflyloxy intermediates,6 which evolve
to vinyl triflates by elimination upon treatment by hindered
pyridines or pyrimidines.7 With this method, almost no regio-
nor stereocontrol could be achieved.8
1,3-Dicarbonyl derivatives, such as 1,3-diketones, ꢀ-ketoal-
dehydes, ꢀ-ketoesters, ꢀ-ketoamides, ꢀ-ketophosphonates and
ꢀ-ketosulfones were efficiently converted to the correspond-
ing Z vinyl triflates with high stereoselectivity. Precoordi-
nation with lithium triflate in dichloromethane and enoliza-
tion with mild bases such as trialkylamines or DBU followed
by trapping with triflic anhydride probably accounted for
such high selectivity, achieved even at 0 °C. This method
offers the first direct route to vinyl triflates from ꢀ-ketoa-
mides, ꢀ-ketophosphonates and ꢀ-ketosulfones.
A third protocol has been developed based on a completely
different process, i.e., the direct addition of triflic acid to
acetylene derivatives (Scheme 2).9
Although the latter method provided a highly regio- and
stereoselective access to vinyl triflates derived from 1,3-
dicarbonyl compounds, it cannot allow to access to vinyl triflates
derived from cyclic or 2-substituted 1,3-dicarbonyl compounds.
In contrast, the two other methods could in principle be applied
to such substituted derivatives. However, these routes were
complicated by the possible formation of mixtures of regioi-
somers and (E)- and (Z)-stereoisomers and most reports seem
specific for particular cases.8,10 A general solution is clearly
needed in this area.
Vinyl triflates are important building blocks in organic
synthesis.1 They are involved in many processes such as
elimination giving alkynes,2 sp2-sp3 coupling reactions with
cuprates3 and mainly sp2-sp and sp2-sp2 coupling reactions,
usually promoted by palladium catalysts.4 Their preparation from
simple ketones commonly relies on two protocols, each involv-
† Present address: Unite´ de chimie organique mole´culaire et macromole´culaire,
Universite´ du Maine, Av. O. Messiaen, Le Mans, France.
Facing such problems in several total syntheses, we designed
new conditions that allow for a convenient and highly regio-
and stereoselective access to Z-vinyl triflates derived from any
1,3-dicarbonyl compounds. Since this method proved quite
general, we detail here our results.
(1) Ritter, K. Synthesis 1993, 735–762.
(2) (a) Hargrove, R. J.; Stang, P. J. J. Org. Chem. 1974, 39, 581–582. (b)
Schwede, W.; Cleve, A.; Ottow, E.; Wiechert, R. Tetrahedron Lett. 1993, 34,
5257–5260. (c) Stang, P. J.; Mangum, M. G.; Fox, D. P.; Haak, P. J. Am. Chem.
Soc. 1974, 96, 4562–4569. (d) Stang, P. J.; Fox, D. P.; Collins, C. J.; Watson,
J. J. Org. Chem. 1978, 43, 364–365. (e) Fleming, I.; Mwaniki, J. M. J. Chem.
Soc., Perkin Trans. 1 1998, 1237–1248.
(3) (a) McMurry, J. E.; Scott, W. J. Tetrahedron Lett. 1980, 21, 4313–4316,
For synthetic applications, see i. a. (b) Kant, J., Jr.; Baker, S. R. Tetrahedron
Lett. 1990, 31, 3389–3392. (c) Pearson, W. H.; Lee, III Y.; Mi, Y.; Stoy, P J.
Org. Chem. 2004, 69, 9109–9122.
(4) (a) Halbes-Letinois, U.; P. Pale, P. Palladium catalyzed synthesis of
Enynes in Leading Edge Organometallic Chemistry Research : Cato, M. A.,
Ed.; Nova Publishers: New York, 2006, pp 93-131. (b) Tsuji, J. Palladium
Reagents and Catalysts; Wiley: New York, 2004. (c) Negishi, E-i.; De Meijere,
A. Handbook of Organopalladium Chemistry for Organic Synthesis; Wiley: New
York, 2002.
(5) Comins, D.; Dehghani, A. Tetrahedron Lett. 1992, 33, 6299–6302.
(6) Wright, M. E.; Pulley, S. R. J. Org. Chem. 1989, 54, 2886–2889.
(7) (a) Stang, P. J.; Treptow, W Synthesis 1980, 283–284. (b) Crich, D.;
Smith, M.; Q. Yao; Picione, J Synthesis 2001, 323–326. (c) Herrera-fernandez,
A.; Martinez-Alvarez, R.; Chioua, M.; Chioua, R.; Almy, J.; Loaiza, O. A.
ARKIVOC 2007, xVi, 58–64.
(8) See, i.a.: Keck, D.; Muller, T.; Bra¨se, S Synlett 2006, 3457–3460.
(9) Vasilyev, A. V.; Walspurger, S.; Chassaing, S.; Pale, P.; Sommer, J. Eur.
J. Org. Chem. 2007, 5740–5748.
10.1021/jo8015049 CCC: $40.75
Published on Web 09/04/2008
2008 American Chemical Society
J. Org. Chem. 2008, 73, 7845–7848 7845

TABLE 1. Screening of Conditions for the Triflation of
2-Formylcycloalkanonesa
TABLE 2. Stereoselective Triflation of 1,3-Dicarbonyl Derivatives
in the Presence of LiOTf, Amines, and Triflic Anhydride at rta
product yieldsb (%)
entry cyclic 1,3-diketones base solvent reagent Z-2
E-2
3c
1
2
3
4
5
6
7
8
9
n)1,1a
n)1,1a
n)1,1a
n)1,1a
n)1,1a
n)1,1a
n)1,1a
n)1,1a
n)2,1b
t-BuLi THF
t-BuLi THF
n-BuLi THF
n-BuLi THF
n-BuLi Et2O
n-BuLi Et2O
n-BuLi CH2Cl2 Tf2O
n-BuLi CH2Cl2 Tf2NPh
n-BuLi CH2Cl2 Tf2O
Tf2O
Tf2NPh
Tf2O
Tf2NPh
Tf2O
Tf2NPh
42
40
40
38
28
27
76
74
75
24
25
26
26
25
24
0
6
7
7
7
17
16
0
0
0
0
0
a Reaction at -78 °C. b Yields of isolated pure products. c 3a proved
to be a extremely sensitive compound, highly prone to degradation.
We first screened conditions on cyclic 1,3-diketones (Table
1), for which only two reports are known.10i,j
Applying conditions described for the selective formation of
Z-2-triflyloxymethylenecyclopentanone,10j we obtained a mix-
ture of regio- and stereoisomers (entry 1). Shifting to a milder
triflating agent did not change the product distribution (entry 2
vs 1). A more aggregate base did not change the product ratio,
regardless of the triflating agent used (entries 3, 4 vs 1, 2).
Running the reaction in a less coordinating solvent surprisingly
increased the amount of the more hindered endocyclic regioi-
somer 3a to the detriment of the Z exocyclic regioisomer Z-2a.
Here again, the nature of the triflating agent did not play any
role (entries 5, 6 vs 3, 4). Decreasing further the coordination
ability of the solvent could favored the Z isomer compared to
the E, but the regioselectivity might be difficult to control as
revealed with the latter examples. Gratifyingly, the reaction
performed in dichloromethane gave the best results, since only
the Z-2-triflyloxymethylenecyclopentanone Z-2a was detected
and isolated whatever the triflating agent (entries 7, 8).
Moreover, these conditions could be applied to the 6-membered
ring analog 1b, leading to the expected Z-2-triflyloxymethyl-
enecyclohexanone Z-2b with the same high regio- and stereo-
selectivity (entry 9). However, applied to noncyclic derivatives,
these conditions led to the Z-stereoisomer in good yields but
the E isomer could be detected and isolated (Z/E 10:1).
These results clearly emphasized the preeminent role of
coordination in such triflation reactions. In order to get even
milder conditions, we envisaged to precoordinate the two
carbonyl groups of the 1,3-diketone moiety. In analogy with
the Evans enolization with boryl triflate,11 we reasoned that such
precoordination should first preorganize the system and second,
increase the polarization of the C-O bonds and acidify the
adjacent proton. A rapid screening revealed that lithium triflate
a Conditions: LiOTf, base in dichloromethane, then Tf2O at 0 °C.
b Yields of isolated
Z products after chromatography; the starting
materials usually accounted for mass balance. c The stereoisomeric ratio
was determined by 1H NMR on the crude mixture. d The product was
instable on silica gel; the yield of crude product was 65%.
in dichloromethane was the perfect choice,12 allowing mild
amine bases to efficiently provide the Z isomer with excellent
stereoselectivity even at 0 °C (Table 2). The premixing time
(10) (a) Hansen, A. L.; Skrydstrup, T. J. Org. Chem. 2005, 70, 5997–6003.
(b) Baxter, J. M.; Steinhuebel, D.; Palucki, M.; Davies, I. W. Org. Lett. 2005,
7, 215–218. (c) Scheiper, B.; Bonnekessel, M.; Krause, H.; Fu¨rstner, A. J. Org.
Chem. 2004, 69, 3943–3949. (d) Bio, M. M.; Leighton, J. L. Org. Lett. 2000, 2,
2905–2907. (e) Romero, D. L.; Manninen, P. R.; Han, F.; Romero, A. G. J.
Org. Chem. 1999, 64, 4980–4984. (f) Kim, H.-O.; Ogbu, C. O.; Nelson, S.;
Kahn, M. Synlett 1998, 1059–1060. (g) Loreto, M. A. Tetrahedron 1997, 53,
15853–15858. (h) Crisp, G. T.; Meyer, A. G. J. Org. Chem. 1992, 57, 6972–
6975. (i) Nakatani, K.; Arai, K.; Yamada, Y.; Terashima, S. Tetrahedron 1992,
48, 3045–3060. (j) Bru¨ckner, R.; Scheuplien, S. W.; Suffert, J. Tetrahedron Lett.
1991, 32, 1449–1452. (k) Saulnier, M. G.; Kadow, J. F.; Tun, M. M.; Langley,
D. R.; Vyas, D. M. J. Am. Chem. Soc. 1989, 111, 8320–8321.
(11) (a) Evans, D. A.; Vogel, E.; Nelson, J. V J. Am. Chem. Soc. 1979, 101,
6120–6121. (b) Evans, D.; Nelson, J. V.; Vogel, E.; Taber, T. R. J. Am. Chem.
Soc. 1981, 103, 3099–3111. (c) Von Horn, D. E.; Masamune, S. Tetrahedron
Lett. 1979, 24, 2229–2232. (d) Hirama, M.; Masamune, S. Tetrahedron Lett.
1979, 24, 2225–2228.
(12) Other lithium salt such as LiCl and Li2CO3 did not give satisfactory
results; 2 equiv of LiOTf was required to achieve high selectivity.
7846 J. Org. Chem. Vol. 73, No. 19, 2008

SCHEME 3
of the expected products could be obtained. These conditions
are therefore limited to triflation reactions.
We then extended the scope of our triflation method to other
1,3-dicarbonyl derivatives. For comparison with other conven-
tional triflation methods, ethyl acetoacetate 6a was submitted
to these conditions. The expected Z-ethyl 3-triflyloxybut-2-
enoate 7a was exclusively produced and isolated in high yield
(entry 13). As with 1,3-diketo derivatives, the use of diisopro-
pylethylamine instead of triethylamine significantly improved
the yield without altering stereocontrol (entry 14 vs 13). It is
also worth noting that this yield is again better than the reported
one (88 vs 62%10g). Another ꢀ-ketoester 6b with a longer chain,
which included a double bond at a suitable position for a
possible cyclization, behaves in the same way, giving the
corresponding triflate Z-7b as a single stereoisomer whatever
the base used (entries 15, 16). The latter result confirmed the
noncationic nature of the reaction mechanism.
The good correlation observed between the lower acidity of
ꢀ-ketoesters and the strength of bases suggested that other
related compounds could also be triflated in these conditions
but shifting from triethylamine to diisopropylethylamine or to
even stronger bases. We thus submitted ꢀ-ketoamides to the
same conditions and screened some bases. With the simple N,N-
diethyl acetylacetamide 8, DBU proved to be the best choice,15
leading to the corresponding triflate 9 in high yield and with
excellent stereoselectivity (entry 17).
Since ꢀ-ketophosphonates and ꢀ-ketosulfones have pKa values
close to those of ꢀ-ketoamides, and due to the analogy between
PdO and SdO bonds with carbonyls, we reasoned that such
compounds might also be triflated in our conditions. Indeed,
ꢀ-ketophosphonate 10 and ꢀ-ketosulfone 12 were efficiently
converted to the corresponding vinyl triflates 11 and 13,
respectively. Interestingly, the same excellent Z stereoselectivity
(>99:1) was observed (entries 18, 19).
In conclusion, we have reported a new and direct route toward
vinyl triflates derived from a large variety of 1,3-dicarbonyl
derivatives, such as 1,3-diketones, ꢀ-ketoaldehydes, ꢀ-ketoesters,
ꢀ-ketoamides, ꢀ-ketophosphonates, and ꢀ-ketosulfones. More-
over, a single regioisomer and stereoisomer was produced with
excellent Z selectivity, except for cyclic ꢀ-ketoaldehydes for
which the selectivity, still high, varied upon ring size.
Further work is now underway to understand this reaction
and its scope.
before adding the triflating agent proved to be critical for
stereoselectivity, since without it, a mixture of stereoisomers
was formed.
Indeed, when 2-formylcyclopentanone 1a was successively
submitted to lithium triflate, triethylamine, and after 20 min to
triflic anhydride in dichloromethane, the expected Z-2-trifly-
loxymethylene cyclopentanone Z-2a was obtained in good yield,
but the other isomer could nevertheless be detected (entry 1).
Interestingly, a slightly stronger base increased both yield and
selectivity (entry 2). The 2-formylcyclohexanone 1b only gave
slightly better results in term of yields, but the selectivity was
twice to four times better (entries 3, 4 vs 1, 2). Since the ring
size seemed to play a role in the coordination due to geometrical
constrains, we submitted the homologous 2-formylcyclohep-
tanone 1c to the same conditions. Indeed, stereoselectivity and
yield were the highest of the series (entries 5, 6 vs 3, 4 vs 1, 2).
It is worth noting that this yield is far better than the reported
one (91 vs 53%13).
Using acyclic 1,3-diketones, these conditions also led to the
formation of the corresponding triflates with excellent stereo-
selectivity (entries 7-12). The simplest acetylacetone 4a gave
the corresponding Z-4-triflyloxypent-3-en-2-one Z-5a with
selectivity as high as 99:1 whatever the base used (entries 7,
8). The stereochemistry was secured by spectroscopic analysis,
especially NOE experiments.
The symmetrical 1,3-diphenyl-1,3-propanedione 4b also gave
a single vinyl triflate, and here also, the more basic amine gave
a higher yield (entry 10 vs 9). However, assigning its stereo-
chemistry was far from obvious since NMR spectra revealed
high shifts for the vinylic proton (7.18 ppm), but without
literature precedent, it was not easy to refer this shift to the
deshielding effect of the carbonyl group common in this series
or to π-effect due to phenyl rings. Fortunately, we were able to
grow crystals from this compound and the X-ray diffraction
pattern unambiguously indicated Z stereochemistry for this
compound (see the Supporting Information).
With the nonsymmetrical 1-phenyl-1,3-butanedione 4c, regio-
and stereoisomers could be expected. Gratifyingly, the preco-
ordination conditions also led to the formation of a single triflate,
with the same yield improvement using stronger amine (entry
12 vs 11). In order to secure its regiochemistry, the vinyl triflate
Z-5c was reduced in the presence of sodium borohydride in
methanol. The corresponding alcohol clearly exhibited in its
1H NMR spectra a singlet for the methyl group and a doublet
for the vinylic and the benzhydryl protons, as expected for the
2,3 regioisomer, clearly different from one would expect from
the other regioisomer (Scheme 3). Crystals could also be
obtained from Z-5c, and X-ray diffraction confirmed its structure
and Z stereochemistry (see the Supporting Information). The
present method thus offers high regioselectivity in the case of
nonsymmetrical 1,3-diketones.
Experimental Section
General Procedure 1 for Triflation. To a 2-formylcycloal-
kanone (1 equiv) in dry solvent (15 mL/mmol) at -78 °C was added
a hexane solution of n-BuLi (1.1 equiv), turning the colorless
solution to a bright yellow mixture. After 10 min of stirring, the
triflating agent (1.1 equiv) was added, discharging the bright color
of the solution. The reaction mixture was further stirred at -78 °C
for 10-30 min depending on the reaction scale. A saturated solution
of ammonium chloride (30 mL/mmol) and dichloromethane (30
mL/mmol) were then added. After the mixture was warmed at room
temperature, the aqueous layer was extracted with dichloromethane
(3 × 30 mL/mmol) and the combined organic layers were dried
over sodium sulfate, filtered, and evaporated.
(Z)-2-Triflyloxymethylenecyclopentanone (Z-2a).10i,j Following
the general procedure 1 using trifluoromethanesulfonic anhydride,
Due to the current focus on the cheaper vinyl tosylates,14 we
also performed some experiments with 4a-c replacing triflic
anhydride by tosyl derivatives. However, no more than 20-30%
(14) (a) Nakatsuji, H.; Ueno, K.; Misaki, T; Tanabe, Y Org. Lett. 2008, 10,
2131–2134. (b) Baxter, J. M.; Steinhuebel, D.; Palucki, M.; Davies, I. W. Org.
Lett. 2005, 7, 215–218.
(13) Bru¨ckner, S.; Abraham, E.; Klotz, P.; Suffert, J. Org. Lett. 2002, 4,
3391–3393.
(15) Simple amines only returned the starting materials.
J. Org. Chem. Vol. 73, No. 19, 2008 7847

2-formylcyclopentanone 1a (0.505 g, 4.5 mmol) gave the Z-
vinyltriflate Z-2a (0.835 g, 3.4 mmol) after 15 min of stirring and
purification on silica gel column chromatography using pentane/
ether (9:1 to 2:1). This compound is extremely sensitive and could
only be stored in dichloromethane solution or better, in benzene
matrix at 0 °C: 1H NMR (300 MHz, CDCl3) δ 6.60 (1H, t, J ) 2.3
Hz), 2.69 (2H, dt, J ) 2.3, 7.2 Hz), 2.40 (2H, t, J ) 7.7 Hz) 2.01
(2H, quint, J ) 7.4 Hz); 13C NMR (75 MHz, CDCl3) δ: 201.75,
133.45, 126.27, 118.43 (q, J ) 321.0 Hz), 39.37, 27.01, 20.33.
General Procedure 2 for Triflation. A solution of ꢀ-dicarbonyl
compound (1 equiv) and lithium trifluoromethanesulfonate (2 equiv)
in dichloromethane (30 mL/mmol) was cooled to 0 °C, and the
base (1.1 equiv) was then added. After 20 min of stirring at 0 °C,
trifluoromethanesulfonic anhydride (1.1 equiv) was added, and the
reaction mixture was further stirred at this temperature upon reaction
completion (followed by TLC). A saturated solution of ammonium
chloride (30 mL/mmol) and dichloromethane (10 mL/mmol) were
then added, and the aqueous phase was extracted twice with
dichloromethane (2 × 30 mL/mmol). The resulting organic layers
were dried over sodium sulfate and evaporated.
39.73, 20.58, 14.31, 12.77; HRMS (ESI) calcd for C9H14O4NF3S
+ Na 312.0488, found 312.0470.
Dimethyl (Z)-2-Triflyloxyprop-1-enephosphonate 11. Follow-
ing the general procedure using 8-diazabicyclo[5.4.0]undec-7-ene,
dimethyl 2-oxopropylphosphonate 10 (0.166 g, 1 mmol) gave the
Z-vinyl triflate 11 after 30 min of stirring. Since the product
decomposed on silica gel, the organic layers were washed with 2
× 30 mL of 5% HCl solution, then with 30 mL of water and 30
mL of brine, dried over Na2SO4, and evaporated to afford 11 as a
1
colorless oil (0.268 g, 90%): H NMR (300 MHz, CDCl3) δ 5.51
(1H, qd, J ) 0.9, 6.4 Hz), 3.77 (6H, d, J ) 11.5 Hz), 2.24 (3H, s);
13C NMR (75 MHz, CDCl3) δ 158.63 (d, J ) 5.6 Hz), 118.34 (q,
J ) 320.1 Hz), 108.92 (d, J ) 191.1 Hz), 52.96 (d, J ) 5.6 Hz),
21.92 (d, J ) 13.0 Hz); 31P NMR (121 MHz, CDCl3) δ 13.13 (s,
1P); HRMS (ESI) calcd for C6H10O6F3PS + Li 305.0042, found
305.0044.
Acknowledgment. We thank the CNRS and the French
Ministry of Research for financial support. S.S. thanks the
French Ministry of Research for a PhD fellowship.
N,N-Diethyl (Z)-3-Triflyloxybut-2-enamide 9. Following gen-
eral procedure 2 using 8-diazabicyclo[5.4.0]undec-7-ene, N,N-
diethyl-3-oxobutanamide 8 (0.158 mL, 1 mmol) gave the Z-vinyl
triflate 9 after 30 min of stirring and purification on silica gel column
chromatography using pentane/ether (2:1 to 1:2) as a colorless oil
Supporting Information Available: Experimental proce-
dures, characterization data, and NMR spectra of compounds
2a-c, 5a-c, 7a,b, 9, 11, and 13; X-ray diffraction structures
of Z-5b and Z-5c and their separate CIF files. This material is
available free of charge via the Internet at http://pubs.acs.org.
1
(0.216 g, 75%): H NMR (300 MHz, CDCl3) δ 5.92 (1H, s), 3.41
(2H, q, J ) 7.1 Hz), 3.33 (2H, q, J ) 7.1 Hz), 2.15 (3H, s), 1.16
(3H, t, J ) 7.1 Hz), 1.13 (3H, t, J ) 7.1 Hz); 13C NMR (75 MHz,
CDCl3) δ: 162.17, 150.29, 118.37 (q, J ) 319.5 Hz), 115.02, 42.58,
JO8015049
7848 J. Org. Chem. Vol. 73, No. 19, 2008