Iodine-mediated Z-selective oxidation of ketones to α,β- unsaturated esters: Synthesis and mechanistic studies

Article abstract of DOI:10.1016/j.tetlet.2005.09.176

The Z-selective oxidation of simple acyclic ketones to Z-2,3-trisubstituted α,β-unsaturated esters is described. Enolates generated under the reaction conditions undergo double iodination followed by a Favorskii-related rearrangement to the unsaturated ester. This reaction represents the first stereoselective one-step transformation of ketones to α,β-unsaturated esters. Mechanistic studies suggest that an electrocyclic reaction governs the Favorskii-related rearrangement.

Full text of DOI:10.1016/j.tetlet.2005.09.176

Tetrahedron
Letters
Tetrahedron Letters 46 (2005) 8289–8292
Iodine-mediated Z-selective oxidation of ketones to
a,b-unsaturated esters: synthesis and mechanistic studies
Michael J. Zacuto* and Dongwei Cai
Department of Process Research, Merck & Co., Inc., PO Box 2000, Rahway, NJ 07065, USA
Received 29 August 2005; revised 22 September 2005; accepted 28 September 2005
Available online 17 October 2005
Abstract—The Z-selective oxidation of simple acyclic ketones to Z-2,3-trisubstituted a,b-unsaturated esters is described. Enolates
generated under the reaction conditions undergo double iodination followed by a Favorskii-related rearrangement to the unsatu-
rated ester. This reaction represents the first stereoselective one-step transformation of ketones to a,b-unsaturated esters. Mechanis-
tic studies suggest that an electrocyclic reaction governs the Favorskii-related rearrangement.
Ó 2005 Elsevier Ltd. All rights reserved.
The conversion of dihaloketones to a,b-unsaturated
esters has received considerable attention and serves as
a valuable alternative to the use of Horner–Wadsworth–
Emmons phosphonoacetate reagents.1 The Favorskii-
related rearrangement of 1,1-dihalo2 and 1,3-dihalo3
ketones to a,b-unsaturated esters has been reported,
including stereospecific reactions that afford the cis-
a,b-unsaturated ester.3a–e This reaction has been applied
primarily to the synthesis of 3-disubstituted acrylates,
including the known stereospecific examples,3a–e while
there are few reports of its use for the synthesis of 2,3-
trisubstituted acrylates2c,d and no selective examples.
One drawback is that these protocols typically involve
a separate step to prepare and isolate the dihaloketone.4
Furthermore, despite the numerous reports, little experi-
mental evidence has been disclosed to either support or
refute the various mechanistic proposals that have been
offered to explain this transformation. A stereoselective
synthesis of Z-2,3-trisubstituted acrylates would be
valuable since these homologues of angelic acid esters
have been studied as medicinal leads5 as well as food
flavorants and fragrance compounds.6 A one-stepproce-
dure which bypasses the isolation of the dihaloketones
would also be an improvement over the existing stereo-
selective protocols. In this letter, we report the one-step
synthesis of Z-2,3-trisubstituted a,b-unsaturated esters
from saturated ketones that proceeds via a diiodoketone
intermediate. We disclose our detailed mechanistic stud-
ies as well and proposed a mechanism for this interesting
reaction.
In the course of our investigations of the iodine-medi-
ated oxidation of carbonyls, we observed a marked dif-
ference in reactivity between cyclic versus acyclic
ketones. Subjection of cyclic six-membered ring ketones
to 1 equiv of iodine and 2 equiv of KOH in MeOH affor-
ded a-hydroxyketals (Eq. 1).7 On the other hand, when
acyclic ketones were subjected to the identical condi-
tions at room temperature a,b-unsaturated esters were
the only identifiable products (Z:E = 3:2) (Eq. 2).
O
MeO OMe
KOH, I2
OH
ð1Þ
ð2Þ
MeOH
89%
O
O
O
R
KOH, I2
MeOH
H
CO2Me
R
R
R
A similar one-steprpocedure using iodine generated
under electrochemical conditions has been reported,2c,d
albeit without selectivity for either olefin isomer. In
addition to improved practicality versus the use of elec-
trochemical cells, the use of elemental iodine allowed for
the opportunity to manipulate reaction conditions in
order to achieve selectivity. Among the numerous vari-
ables examined, we found that order of addition was
crucial for obtaining high selectivity. Addition of
KOH/MeOH solution to a cold solution of ketone and
Keywords: Iodine; Oxidation; Acrylates.
*
Corresponding author. Tel.: +1 732 5940363; fax: +1 732 5945170;
e-mail: michael_zacuto@merck.com
0040-4039/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tetlet.2005.09.176

8290
M. J. Zacuto, D. Cai / Tetrahedron Letters 46 (2005) 8289–8292
I2 in MeOH allowed for the stereoselective preparation
of Z-acrylates. By contrast, the reverse order of addition
resulted in lower selectivity (Z:E = 7:1). The optimized
conditions involved the slow addition of KOH
(4.8 equiv) in MeOH to a À5 °C MeOH solution con-
taining ketone and I2 (2.2 equiv). This procedure was
applied to a series of simple symmetric ketones and
the results are presented in Table 1.8
respect to the products formed when unsymmetrical
ketones (7, R1 5 R2) are employed. The lack of
selectivity seemingly follows from the previously pro-
posed mechanism,2c,d whereby a-iodoketone 8 under-
goes further iodination to a,a-diiodoketone 9. A
presumed Favorskii rearrangement via a-iodocyclo-
propanone 10 to the saturated b-iodoester 11 is followed
by elimination of HI to afford 12. Accordingly, a selec-
tive initial enolization (selective production of 8, where
R1 5 R2) would result in selective product formation.
With this in mind, we subjected the known iodoketone
139 to modified reaction conditions. Treatment of 13
with 1.1 equiv of I2 and 3.4 equiv of KOH in MeOH
afforded a 1:1 mixture of regioisomeric acrylates 14
and 15 (Z:E = 25:1).
Unbranched symmetric ketones are excellent substrates
for the oxidation reaction. Branching at the b-position
of the ketone leads to a lower Z:E ratio (Table 1, entry
5), as does conducting the reaction at room temperature
(Z:E = 11:1 in the case of 3-pentanone). While water is
present in the reaction mixture (KOH serves as a conve-
nient precursor to KOMe), saponification of the product
esters was not observed under these reaction conditions.
O
O
O
I2
MeO
I2
MeO
I
I
I
We also considered extending this reaction to unsym-
metrical ketones. Previous electrochemical studies,2c,d
however, demonstrated a lack of regioselectivity with
MeOH
MeOH
R1 R2
7
R1 R2
8
R1 R2
9
MeO
O
R2
R2
MeO
MeO
Table 1. Synthesis of a,b-unsaturated esters from ketones
I
I
CO2Me
CO2Me
MeOH
MeOH
I2
R
R1
R1
R1 R2
10
O
KOH
12
O
11
H
R
R
CO2Me
MeOH
R
-5 to 0 oC
I2
KOH
Me
n
-Pr
Entry
1
R
Product
Z:Ea
Yieldb (%)
60d
I
H
+
H
CO2Me
CO2Me
MeOH
Me
n-Pr
n-Pr
Me
CO2Me
Me
(1:1)
Me
20:1c
H
H
13
14
15
Me
1
The conversion of 13 to 14 and 15 effectively ruled out
the possibility of selective product formation from
unsymmetrical ketones. Furthermore, this experiment
suggested that a different mechanism may be operating
and would explain the observed product distribution.
Though the synthetic utility of the reaction is relegated
to the synthesis of angelate homologues, we became
intrigued by the mechanism of the overall transformation
and turned our attention towards detailed mechanistic
studies. We also hoped to shed light on the various
mechanistic proposals in the literature for the conver-
sion of dihaloketones to a,b-unsaturated esters.
Et
2
3
4
Et
25:1
25:1
25:1
72
80
83
CO2Me
2
Et
n
-Pr
H
H
n-Pr
CO2Me
3
n
-Pr
n-Bu
n-Bu
CO2Me
4
n-Bu
We began by verifying the importance of a diiodoketone
(9 or an isomer) as an intermediate under our reaction
conditions. Subjection of 5-nonanone (16) to only half
of the reagents (2.4 equiv KOH, 1.1 equiv of I2) under
otherwise identical reaction conditions led to the forma-
tion of a 1:1 mixture of 3 and unreacted 16. By analogy
to the haloform reaction, this suggested that an initially
i
-Pr
H
5
6
i-Pr
9:1
75
75
CO2Me
i
-Pr
5
Ph
H
CH2Ph
20:1
CO2Me
6
Ph
O
O
n
-Pr
KOH, I2
MeOH
a Ratio determined by 1H NMR.
b Isolated yield.
+
H
CO2Me
n
-Pr n-Pr
n
-Pr n-Pr
n
-Pr
c Ratio determined by HPLC.
d LC assay yield.
(1:1)
3
16
16

M. J. Zacuto, D. Cai / Tetrahedron Letters 46 (2005) 8289–8292
8291
formed a-iodoketone underwent enolization and iodina-
tion faster than the starting ketone.
ment, where subjection of 5-nonanone (16) to the
reaction conditions (KOD, I2, CD3OD) afforded the
acrylate products 24 and 25 with only 20% D incorpora-
tion (80% H) at the b-position. If operative, the classical
Favorskii pathway would have afforded both saturated
iodo-esters 27 and 28, followed by elimination of HI.13
If 27 were formed exclusively, then the product would
have 100% D incorporation at the b-position of the
product acrylate. If 28 were formed exclusively, then
the amount of D incorporation at the b-position of the
product would depend upon the kinetic isotope effect,
but should be >50% since normal primary kinetic iso-
tope effect are observed in E2 elimination reactions.
The observation of only 20% D incorporation in the
product suggested that neither 27 nor 28 were precursors
to the product acrylate.
Since both a,a-dihaloketones 172 and a,a0-dihalo-
ketones 183 are reported to undergo alkoxide-promoted
conversion to a,b-unsaturated esters via an intermediate
a-halocyclopropanone (19), we sought to distinguish
between the two possible intermediates in the present
study. Due to their symmetry, the substrates examined
in Table 1 could not provide information to differentiate
these two pathways. Thus the reactivity of iodoketone
13 was probed further.
O
R
O
MeO
MeO
R
R
H
R
R
CO2Me
MeOH
MeOH
X
X
R
X
X
17
18
O
n-Pr
n-Pr
I2, KOD
CD3OD
76%
H
D
O
CO2CD3
+
CO2CD3
25
n
-Pr
n
-Pr
n-Pr
n-Pr
R
H
(4:1)
16
24
19
R
X
Compound 13 was treated with 0.8 equiv of I2 and
0.7 equiv of KOH in MeOH in an NMR experiment.
The newly observed chemical shifts were consistent with
the formation of a,a-diiodoketone 20.10 Further treat-
ment with 1.0 equiv of KOH led to the formation of
small amounts of acrylates 14 and 15, along with a
new compound whose NMR data was consistent with
a,a0-diiodoketone 21.11 Based on this result, we propose
that an a,a0-diiodoketone (21) is the actual diiodoketone
species which undergoes Favorskii-type rearrangement
to the product a,b-unsaturated esters. In the case of
unsymmetrical ketones, two intermediate iodocyclo-
propanones 22 and 23 can form, which likely accounts
for the unselective formation of both 14 and 15.
O
CO2CD3
R
CO2CD3
n-Pr
H
R
D
R
+
R
I
D
I
n-Pr
I
H
27
28
26
Since saturated iodo-esters (e.g., 27 and 28) were not
likely intermediates, it seemed plausible that the iodo-
cyclopropanone was converted directly to the ester prod-
uct. A proposed mechanism that accounts for this is
an electrocyclic reaction. The transformation of halo-
cyclopropanone intermediates into a,b-unsaturated carbo-
nyls has been explained by the disrotatory electrocyclic
ring opening of halocyclopropane to the allyl cat-
ion.3a,b,14a In this scenario, enolization of the intermedi-
ate 18 would lead to the presumed cis-cyclopropanone
30,12,14a perhaps via disrotatory electro-cyclic ring
closure of zwitterion 29.15 Though we have not been
able to observe the cyclopropanone intermediate, the
stereochemistry of the observed acrylate product implies
the cis configuration according to the Woodward–Hoff-
mann rules.12 Additionally, ConiaÕs verification of these
rules in the stereoselective conversion of analogous cis-
chlorocyclopropanols to Z-enals lends support to this
O
O
O
I2
Me
n-Pr
KOMe Me
KOMe
Me
MeOH
I
I
I
n-Pr
n
-Pr
I
I
13
20
21
O
O
Me
n-Pr
14 + 15
+
I
n
-Pr
I
Me
23
22
O
O
R
i. MeO-
ii. - I
H
R
R
R
O
R
We next turned our attention towards elucidating the
conversion of iodocyclopropanone to a,b-unsaturated
ester. Two mechanisms have been proposed in the litera-
ture. The first involves a classical Favorskii rearrange-
ment (i.e., 10 ! 11 ! 12),2,3h while the second involves
an electrocyclic reaction.3a,b,12
I
I
I
I
R
30
18
R
29
MeO-
R
I
OMe
R
O
H
The present transformation likely does not follow from
the classical Favorskii mechanism. We base this conclu-
sion on the results from a deuterium labelling experi-
CO2Me
H
OMe
O
R
R
R
31

8292
M. J. Zacuto, D. Cai / Tetrahedron Letters 46 (2005) 8289–8292
speculated stereochemistry.14a Addition of methoxide to
30 would afford 31, which can undergo alkoxide-pro-
moted disrotatory electrocyclic ring opening with con-
certed expulsion of iodide3a,b,12 to afford the Z-
acrylate. Given the available data, this seems to be the
most reasonable mechanism.
suitable for further use, but could be purified by silica
gel chromatography. Data for 1: 1H NMR, 13C NMR, MS
and HPLC agree with data obtained from commercially
available material. Data for 2: 1H NMR (CDCl3,
400 MHz): d 5.83 (t, J = 7.6 Hz, 1H), 3.73 (s, 3H), 2.45–
2.35 (m, 2H), 2.26 (q, J = 7.6 Hz, 2H), 1.02 (t, J = 7.6 Hz,
3H), 1.00 (t, J = 7.6 Hz, 3H); 13C NMR (CDCl3,
400 MHz): d 168.6, 142.2, 132.8, 51.0, 27.4, 22.9, 13.9,
1
In summary, we have described the iodine-mediated
one-stepoxidation of symmetric ketones to Z-a,b-unsat-
urated esters. This method offers rapid access to valu-
able acrylates from easily accessible and inexpensive
starting materials, with no organic waste products.
Mechanistic studies suggest that an a,a0-diiodoketone
undergoes Favorskii-related rearrangement, via an elec-
trocyclic reaction, to afford a,b-unsaturated ester.
13.6. Data for 3: H NMR (CDCl3, 400 MHz): d 5.84 (t,
J = 7.2 Hz, 1H), 3.73 (s, 3H), 2.37 (td, J1 = 7.6 Hz,
J2 = 7.6 Hz, 2H), 2.22 (t, J = 7.6 Hz, 2H), 1.48–1.37 (m,
4H), 0.92 (t, J = 7.2 Hz, 3H), 0.89 (t, J = 7.2 Hz, 3H); 13
C
NMR (CDCl3, 400 MHz): d 168.7, 141.6, 131.8, 50.9, 36.5,
1
31.5, 22.6, 22.1, 13.6, 13.4. Data for 4: H NMR (CDCl3,
400 MHz): d 5.84 (t, J = 7.4 Hz, 1H), 3.73 (s, 3H), 2.39 (td,
J1 = 7.2 Hz, J2 = 7.2 Hz, 2H), 2.23 (t, J = 7.6 Hz, 2H),
1.43–1.25 (m, 8H), 0.99 (t, J = 7.2 Hz, 3H), 0.89 (t,
J = 7.2 Hz, 3H); 13C NMR (CDCl3, 400 MHz): d 168.8,
141.7, 132.0, 51.1, 34.3, 31.7, 31.3, 29.3, 22.4, 22.2, 13.9,
13.9. Data for 5: 1H NMR (CDCl3, 400 MHz): d 5.46 (dd,
J1 = 9.6 Hz, J2 = 1.2 Hz, 1H), 3.74 (s, 3H), 2.88–2.78 (m,
1H), 2.69–2.57 (m, 1H), 1.04 (t, J = 6.8 Hz, 3H), 0.98 (t,
J = 6.5 Hz, 3H); 13C NMR (CDCl3, 400 MHz): d 169.5,
141.8, 136.6, 50.9, 31.2, 28.5, 22.7, 21.7. Data for 6: 1H
NMR (CDCl3, 400 MHz): d 7.38–7.20 (m, 10H), 6.11 (t,
J = 7.6 Hz, 1H), 3.85 (d, J = 7.6 Hz, 2H), 3.72 (s, 3H),
3.63(s, 2H); 13C NMR (CDCl3, 400 MHz): d 167.8, 141.6,
140.0, 139.2, 131.6, 128.7, 128.5, 128.5, 128.3, 126.2, 126.2,
51.3, 40.4, 35.8.
References and notes
1. Wadsworth, W. S., Jr. Org. React. 1977, 25, 73; For
example, the use of phosphonoacetate reagents for the
synthesis of Z-2,3-trisubstituted acrylates is hampered by
the susceptibility of the products to isomerization under
the reaction conditions; see Wadsworth, W. S., Jr.;
Emmons, W. D. Org. Synth. 1965, 45, 44.
2. (a) Morimoto, T.; Sekiya, M. Chem. Pharm. Bull. 1982,
30, 3513; (b) Kimpe, N. D.; Stanoeva, E.; Boeykens, M.
Synthesis 1994, 427; For a proposed a,a-diiodo interme-
diate see (c) Barba, F.; Elinson, M. N.; Escudero, J.;
Guirado, M.; Feducovich, S. K. Electrochim. Acta 1998,
43, 973; (d) Elinson, M. N.; Feducovich, S. K.; Zaimov-
skaya, T. A.; Dorofeev, A. S.; Vereshchagin, A. N.;
Mikishin, G. I. Russ. Chem. Bull., Int. Ed. 2003, 52, 998.
3. (a) Engler, T. A.; Falter, W. Tetrahedron Lett. 1986, 27,
4115; (b) Engler, T. A.; Falter, W. Tetrahedron Lett. 1986,
27, 4119; (c) Rappe, C. Org. Synth. 1973, 53, 123; (d)
Kennedy, J.; McCorkindale, N. J.; Raphael, R. A.; Scott,
W. T.; Zwanenburg, B. Proc. Chem. Soc. 1964, 168; (e)
Rappe, C. Acta. Chem. Scand. 1963, 17, 2766; (f) Wagner,
R. B.; Moore, J. A. J. Am. Chem. Soc. 1950, 72, 3655; (g)
Abad, A.; Arno, M.; Pedro, J. R.; Seoane, E. Chem. Ind.
1981, 5, 157; (h) Krabbenhoft, H. O. J. Org. Chem. 1979,
44, 4285.
9. Hojo, H.; Harada, H.; Ito, H.; Hosomi, A. J. Am. Chem.
Soc. 1997, 119, 5459.
10. 1H and 13C NMR data agreed with data for known a,a-
diiodoketones. For example, see: Heasley, V. L.; Shell-
hamer, D. F.; Chappell, A. E.; Cox, J. M.; Hill, D. J.;
McGovern, S. L.; Eden, C. C.; Kissel, C. L. J. Org. Chem.
1998, 63, 4433.
11. The 1H and 13C data agrees well with known data for 2,4-
diiodo-3-pentanone: see Montanˇa, A. M.; Grima, P. M.
Synth. Commun. 2003, 33, 265–279; For a similar iso-
merization, see: Rappe, C. Acta Chem. Scand. 1963, 17,
2140.
12. (a) Hoffmann, R.; Woodward, R. B. Accounts Chem. Res.
1968, 1, 17, and references cited therein; (b) DePuy, C. H.
Accounts Chem. Res. 1968, 1, 33.
13. The iodocyclopropanone intermediate 26 is presumed to
contain a proton at the a-carbon, as isotopic scrambling at
the a-carbon of the ketone should be minimal since the
procedure involves the slow addition of base to a mixture
of the ketone and I2. This assumption was affirmed when
the reaction was run to 50% conversion and NMR data
indicated only 20% D incorporation into the unreacted
ketone.
4. For an exception, see Ref. 2c,d. No selectivity is observed
under these conditions.
5. (a) Palaty, J.; Abbott, F. S. J. Med. Chem. 1995, 38, 3398;
(b) Okada, K.; Kiyoka, F.; Nakanishi, E.; Hirano, M.;
Ono, J.; Matsuo, N.; Matsui, M. Agric. Biol. Chem. 1980,
44, 2595.
6. Japanese Patent # JP60028950; 1985.
7. Zacuto, M. J.; Cai, D. Tetrahedron Lett. 2005, 46, 447.
8. General experimental procedure for the synthesis of 1–5:
Ketone (5.0 mmol) was added to a solution of I2 (2.8 g,
11 mmol) in MeOH (20 mL), then cooled to À5 °C. A
solution of KOH (1.6 g of A.C.S. 85% KOH assay pellets)
in MeOH (18 mL) was then added over 25 min. The
solution was then allowed to warm gradually to room
temperature, and then stirred for an additional 1 h. The
solution was then concentrated, and the resulting slurry
was diluted with Pentane and filtered. Concentration
afforded the crude a,b-unsaturated ester which was
14. For related studies on isolated cyclopropane systems, see:
(a) Conia, J.-M.; Blanco, L. New J. Chem. 1983, 7, 399; (b)
Slougui, N.; Rousseau, G. Synth. Commun. 1982, 12, 401;
(c) Slougui, N.; Rousseau, G. Tetrahedron 1985, 41, 2643.
15. For a review of the mechanism of the Favorskii rear-
rangement, see: (a) Baretta, A.; Waegell, B. In Reactive
Intermediates; Abramovich, R. A., Ed.; Plenum: New
York, 1982; Vol. 2, Chapter 6. The stereochemistry of 29 is
speculated, but implied by the presumed stereochemistry
of 30 according to the Woodward–Hoffmann rules (see,
Refs. 12, 14a).