Regioselective cis,vic-dihydroxylation of α,β,γ,δ- unsaturated carboxylic esters: Enhanced γ,δ-selectivity by employing trifluoroethyl or hexafluoroisopropyl esters

Article abstract of DOI:10.1055/s-2006-951473

The regioselectivity of Sharpless asymmetric dihydroxylation (AD) of α,β,γ,δ-unsaturated carboxylic esters was studied as a function of α-, β-, and δ-substituents and for fluorine-free versus fluorinated esters. The latter showed increased or complete γ,δ-selectivities: the hexafluoroisopropyl ester being superior to the trifluoroethyl ester. Olefinations of α,β-unsaturated aldehydes with phosphorus ylide 36 or phosphonate anion 41 provided α,β, γ,δ-unsaturated trifluoroethyl esters, leading inter alia to complete trans selectivity and to 31 with 94% E selectivity, respectively. Georg Thieme Verlag Stuttgart.

Full text of DOI:10.1055/s-2006-951473

LETTER
2641
Regioselective cis,vic-Dihydroxylation of a,b,g,d-Unsaturated Carboxylic
Esters: Enhanced g,d-Selectivity by Employing Trifluoroethyl or
Hexafluoroisopropyl Esters
R
egioselec
o
tive cis
,
vic-D
a
ihydroxyl
c
ation of
a
,
b
h
,
g
,
d
-Unsatur
i
ated Ca
m
rboxylic Esters Schmidt-Leithoff, Reinhard Brückner*
Institut für Organische Chemie und Biochemie, Albert-Ludwigs-Universität Freiburg, Albertstr. 21, 79104 Freiburg, Germany
Fax +49(761)2036100; E-mail: reinhard.brueckner@organik.chemie.uni-freiburg.de
Received 9 June 2006
As described in the following we were able to shift the
regioselectivity of the mono-AD of various a,b,g,d-unsat-
urated carboxylic esters from a given g,d:a,b ratio towards
g,d or entirely to g,d by replacing methyl or ethyl esters by
Abstract: The regioselectivity of Sharpless asymmetric dihydroxy-
lation (AD) of a,b,g,d-unsaturated carboxylic esters was studied as
a function of a-, b-, and d-substituents and for fluorine-free versus
fluorinated esters. The latter showed increased or complete g,d-se-
lectivities: the hexafluoroisopropyl ester being superior to the tri- trifluoroethyl esters or in one case by the hexafluoroiso-
fluoroethyl ester. Olefinations of a,b-unsaturated aldehydes with
phosphorus ylide 36 or phosphonate anion 41 provided a,b,g,d-un-
saturated trifluoroethyl esters, leading inter alia to complete trans
selectivity and to 31 with 94% E selectivity, respectively.
propyl ester. The dienoic esters of our substrates were the
reportedly difficult cases 13 and 33 or dictated by a syn-
thetic objective (5/76) or contained d-alkynyl substituents,
a d-acetyl group or an acetal.
Key words: asymmetric dihydroxylation, hexafluoroisopropyl es-
ter, regioselectivity, trifluoroethyl esters, a,b,g,d-unsaturated esters
β
δ
γ
α
CO2Et
Ph
1
The Sharpless catalytic asymmetric dihydroxylation (AD)
is one of the key tools of organic synthesis.1 While
applicable to many monoolefins relatively independently
from their substitution pattern, AD reactions of dienes
may be problematic with respect to achieving chemo- and
regioselectivity. It is common for conjugated dienes,
though, that their ADs can be stopped at the stage of
mono-ADs.2–4 This is due to the increase of steric hin-
drance and the decrease of electron density once the first
C=C bond has reacted. In contrast, AD reactions of 1,3-
dienes often lack regiocontrol. For instance, the dienoic
ester 1 can be monohydroxylated but undergoes a,b-AD
(‘proximal functionalization’, disfavored) along with g,d-
AD (‘distal functionalization’, favored; Scheme 1).3 Like-
wise, distal rather than proximal functionalization is
preferred in the mono-AD of 2,4,6-octatrienoates.4 The
conclusion at that time, that mono-ADs of polyenes occur
preferentially at their most electron-rich double bond,4
disagrees with the regioselectivity of the mono-AD 3 → 2
+ iso-2 observed later.3 The ester-substituted C=C bond of
3 would be regarded as more electron-rich than the ke-
tone-substituted C=C bond; nonetheless, the former reacts
faster. Neither are the g,d-:a,b-AD ratios in a series of 5-
phenyl-2,4-pentadienoic acid derivatives accommodated
by the reactivity order ‘electron-rich C=C bond reacts
prior to electron-deficient C=C bond’: These ratios are 2:1
for the dihydroxylation of the tert-butyl ester in the
presence of (DHQ)2PHAL, 6.5:1 for the ethyl ester, and
>20:1 for the Weinreb amide.5
a (73%)
Ph
OH
OH
α
γ
CO2Et
CO2Et
δ
+
β
Ph
OH
OH
2
87:13
iso-2
CO2Et
β
δ
γ
α
O
3
b (59%)
O
OH
OH
α
γ
CO2Et
CO2Et
δ
+
β
OH
O
OH
4
60:40
iso-4
Scheme 1 ADs of ethyl esters of a,b,g,d-unsaturated carboxylic
acids exhibiting incomplete regioselectivities (ref. 3). Reagents and
conditions: a) Modified AD-mix™ a [namely K2OsO2(OH)4 (1
mol%), (DHQ)2PHAL (1 mol%)], 0 °C; b) modified AD-mix™ a
[K2OsO2(OH)4 (1 mol%), (DHQ)2PHAL (2 mol%), K3Fe(CN)6 (2
equiv), K2CO3 (2 equiv), t-BuOH–t-BuOMe–H2O (1:1:2)], 0 °C, no
reductive workup.
The fluorinated esters of our study turned out to be con-
siderably less reactive than their fluorine-free counter-
parts. Therefore, we employed higher-than-usual amounts
of osmate (1–5 mol%) and enantiopure ligand (5–10
mol%) in their ADs. This measure is reminiscent of, but
clearly preferable to, the use of 11 mol% of OsO4 and 10
mol% of the ligand in the mentioned mono-AD of the di-
enoic Weinreb amide.5
Table 1 shows pairs of mono-ADs of 5-alkynylated 2,4-
dienoic esters. Methyl 2,4-dienoate 5 with the tert-
BuMe2SiC≡C-substituent at C-5 showed a 85:15 prefer-
ence for g,d- versus a,b-dihydroxylation. The analogous
SYNLETT 2006, No. 16, pp 2641–2645
Advanced online publication: 22.09.2006
DOI: 10.1055/s-2006-951473; Art ID: G18306ST
© Georg Thieme Verlag Stuttgart · New York
0
4
.1
0
.2
0
0
6

2642
J. Schmidt-Leithoff, R. Brückner
LETTER
trifluoroethyl ester 7 gave only g,d-dihydroxylation prod-
uct 8.6 Similarly, methyl 2,4-dienoate 9 with the PhC≡C-
substituent at C-5 gave a 70:30 mixture of the g,d- and
a,b-dihydroxylation products 108 and iso-10 whereas the
analogous trifluoroethyl ester 11 delivered the g,d-dihy-
droxylation product 12 selectively. In these AD reactions
we used five times as much9 osmate (namely 1 mol%) and
AD-mix™ a ligand (namely 5 mol%) as routinely
recommended10 in order to achieve ca. 70% yields within
ca. 24 hours.11 Obviously, alkyne substituents slow down
Os(VIII)-mediated cis,vic-dihydroxylations.14,15
Table 2 Regiocontrolled Dihydroxylations of d-Alkynyl a,b,g,d-
Unsaturated Esters Containing Trifluoroethoxy versus Ethoxy
Moieties
δ
CO2R2
γ
R1
β
α
1, 13, 3, 15, 17, 19
modified AD-mix™ α
OH
OH
CO2R2
CO2R2
γ
α
R1
R1
+
β
δ
OH
OH
2, 14, 4, 16, 18, 20
iso-2, iso-4, iso-18
Table 1 Regiocontrolled Dihydroxylations of d-Alkynyl-b-methyl
a,b,g,d-Unsaturated Esters with Trifluoroethoxy versus Alkoxy
Moietiesa
Substrate
R1
Ph
R2
Et
Product(s) Yield (%) g,d-:a,b-attack
1a
2/iso-2b
14
71
63
92:8
100:0
13a
Tfec
δ
CO2R2
γ
β
α
3d
Tfec
Tfec
4/iso-4e
16
55
44
73:27
100:0
R1
5, 7, 9, 11
15d
O
‘improved’ AD-mix™ α
OH
OH
17f
19f
Et
18/iso-18b
45
39
65:35
100:0
CO2R2
CO2R2
O
O
γ
α
δ
Tfec
20
+
β
R1
OH
iso-6, iso-10
R1
OH
6, 8, 10, 12
a Conditions same as in footnote a of Table 1.
b Inseparable by flash chromatography7 on silica gel.
c Tfe = trifluoroethyl.
Substrate R1
R2
Product(s) Yield (%) g,d-:a,b-attack
d Conditions same as in footnote a of Table 1 but less K3Fe(CN)6 (2.0
equiv), less K2CO3 (2.0 equiv), 0 °C, 4 d, no workup with Na2S2O3.
e Compound 4 was isolated pure, iso-4 in a mixture with MeSO2NH2.
f Conditions same as in footnote a of Table 1 but 48 h.
56
Me
6/iso-6b
65
72
85:15
100:0
TBSc
76
Tfed
8
9
11
Me
10/iso-10b
79
65
70:30
100:0
Ph
Tfed
12
The last formula line of Table 4 illustrates how ADs of
conjugated dienoic esters may be improved from an im-
perfect to a virtually perfect g,d-:a,b-regioselectivity
based on the ‘fluorine-containing versus fluorine-free
a K2Os(OH)4O2 (1 mol%), (DHQ)2PHAL (5 mol%), K3Fe(CN)6
(3.0 equiv), K2CO3 (3.0 equiv), MeSO2NH2 (1.0 equiv), t-BuOH–H2O
(1:1), 0 °C, 24 h; workup with Na2S2O3.
b Inseparable by flash chromatography7 on silica gel.
c TBS = tert-butyldimethylsilyl.
d Tfe = trifluoroethyl.
Table 3 Regioselectivity of Dihydroxylations of d-Alkynyl a,b,g,d-
Unsaturated Esters Containing Trifluoroethoxy versus Ethoxy
Moieties
Table 2 documents how our ‘fluorine-containing versus
fluorine-free ester concept’ allows to convey complete
g,d-regioselectivity to the mono-AD of the a,b,g,d-unsat-
urated esters 13 versus 1 (Ph being the d-substituent), 15
versus 3 (acetyl being the d-substituent), and 19 versus 17
(a ketal being the d-substituent). While the ethyl esters
reacted with 65:3516–92:816 g,d- versus a,b-preferences,
their trifluoroethyl analogues gave 100:0 ratios.
δ
CO2R2
γ
β
α
R1
21, 23, 25, 27
modified AD-mix™ α
OH
OH
CO2R2
CO2R2
γ
α
β
δ
+
R1
OH
R1
OH
Some trifluoroethyl 2,4-dienoates were not cis,vic-dihy-
droxylated with complete g,d-selectivity (Tables 3 and 4).
This concerned scaffolds where the fluorine-free sub-
strates react at C-a and C-b preferentially (21, 25) or al-
most as much as at C-g and C-d (29). In these cases, the
fluorinated counterparts 23, 27, and 31 allowed to over-
come this bias qualitatively but not completely. Accord-
ingly, the latter substrates gave 50:50, 69:31, and 92:8
mixtures of the g,d- and a,b-diol isomers 24/iso-24, 28/
iso-28, and 32/iso-32, respectively. The 32% yields both
of 24/iso-24 and 28/iso-28 could not be raised to the 73%
which was the yield of the 32/iso-32 mixture, not even
when employing, as we actually did, 5 mol% of
K2Os(OH)4O2 and 10 mol% of (DHQ)2PHAL.
22, 24, 26, 28
iso-22, iso-24, iso-26, iso-28
Substrate R1
R2
Productsa
Yield (%) g,d-:a,b-attack
2b
Et
22/iso-22
24/iso-24
59
32
12:88
50:50
TBSc
23d
Tfee
25b
27d
Et
26/iso-26
28/iso-28
53
32
23:77
69:31
Ph
Tfee
a Inseparable by flash chromatography7 on silica gel.
b Conditions same as in footnote a of Table 1 but more K2Os(OH)4O2
(5 mol%), more (DHQ)2PHAL (10 mol%), 48 h.
c TBS = tert-butyldimethylsilyl.
d Conditions same as in footnote b but also NaHCO3 (3.0 equiv).
e Tfe = trifluoroethyl.
Synlett 2006, No. 16, 2641–2645 © Thieme Stuttgart · New York

LETTER
Regioselective cis,vic-Dihydroxylation of a,b,g,d-Unsaturated Carboxylic Esters
2643
ester concept’: by dihydroxylating the respective hexa-
fluoroisopropyl ester (specifically 3317) – on grounds of
expecting its Ca=Cb bond to be even electron-poorer than
the Ca=Cb bond in the corresponding trifluoroethyl ester
(specifically 31). Under the forcing AD conditions which
had provided the 92:8 mixture of g,d- and a,b-dihydroxy
trifluoroethyl esters 32 and iso-32, we now obtained the
g,d-dihydroxy hexafluoroisopropyl ester 3418 as a single
isomer (40% yield).
a, b
c
Br
CO2Tfe
3520
O
Ph3P
CO2Tfe
Ph
(MeO)2P
CO2Tfe
3822
O
36
37
e
d
CO2Tfe
Ph
13
Table 4 Regiocontrolled Dihydroxylation of an a-Methyl-d-
alkynyl a,b,g,d-Unsaturated Ester Containing Differently Fluorinated
Scheme 2 Stereoselective synthesis of an a,b,g,d-unsaturated
trifluoroethyl (Tfe) ester. Reagents and conditions: a) PPh3 (1.0
equiv), toluene, r.t., 20 h, 100% (ref. 23 mentions no yield); b) NaOH
(2 M, 1.0 equiv), H2O, 0 °C, 100% (ref. 23 mentions no yield); c)
(MeO)3P (1.3 equiv), 60–70 °C, 1 h; 4 h, 90 °C, 98%; d) 36 (3.0
equiv), CH2Cl2, r.t., 48 h, 100%, 2E,4E:2E,4Z (98.5:1.5) mixture;
e) 38 (1.0 equiv), NaH (1.0 equiv), THF, 0 °C, 10 min; addition of 37,
–78 °C, THF, 90 min, 65%, 2E,4E:2Z,4E (96:4) mixture.
Alkoxy Moieties
δ
γ
α
CO2R
β
Ph
29, 31, 33
modified AD-mix™ α
OH
OH
γ
α
CO2R
OH
CO2R
+
β
δ
Ph
Ph
OH
30, 32, 34
O
iso-30, iso-32
(MeO)2P
CO2Tfe
Br
CO2H
Br
CO2Tfe
a
b
Substrate
29a
R
Products
30/iso-30b
32/iso-32b
34
Yield (%) g,d-:a,b-attack
39
42
40
41
Et
70
73
40
56:44
92:8
31c
Tfed
Hfipf
f
c–e
O
Ph
Ph
33e
100:0
43
a Conditions same as in footnote a of Table 1.
3
5
2
4
CO2Tfe
b Separable by flash chromatography7 on silica gel.
c Conditions same as in footnote a of Table 1 but more K2Os(OH)4O2
(5 mol%), more (DHQ)2PHAL (10 mol%), 48 h.
d Tfe = trifluoroethyl.
Ph
31
Scheme 3 Stereoselective synthesis of an a-methylated a,b,g,d-un-
saturated trifluoroethyl (Tfe) ester. Reagents and conditions: a)
F3CCH2OH (3.0 equiv), H2SO4 (cat.), benzene, reflux, 3 d, 78%; b)
(MeO)3P (1.3 equiv), 180 °C, 4 h, 44%; c) CuCl (5 mol%),
NH2OH·HCl (cat.), n-BuNH2 (30% in H2O), 0 °C; then 42, 3-bromo-
2-propyn-1-ol (1.0 equiv), 0 °C → r.t., 2 h, 90%; d) LiAlH4 (1.5
equiv), Et2O, 0 °C → r.t., 3.5 h, 96%; e) MnO2 (30 equiv), CH2Cl2,
r.t., 4 h, 82%; f) 41 (1.2 equiv), n-BuLi (1.4 equiv), THF, -78 °C, 10
min; then 43, 30 min; 0 °C, 2 h, 73% of a 2E,4E:2Z,4E:2E,4Z
(90:6.4:3.6) mixture.
e Conditions same as in footnote c but also NaHCO3 (3.0 equiv).
f Hfip = hexafluoroisopropyl.
The a,b,g,d-unsaturated esters of the present study were
prepared by Negishi, Heck or Stille couplings or by a Wit-
tig or Horner–Wadsworth–Emmons (HWE) reaction.19
Scheme 2 illustrates Wittig reagent 36 in a completely
trans-selective chain-extension of cinnamic aldehyde
(37); it gave the trifluoroethyl dienoate 13 in quantitative
yield.24,25 A 65% yield of the same dienoate was obtained
with 96:4 trans selectivity by an HWE reaction between
cinnamic aldehyde (37) and deprotonated (NaH) trifluo-
roethyl phosphonylacetate 38.22
Acknowledgment
We are indebted to the Fonds der Chemischen Industrie for financi-
al support. We thank Tatjana Stricker for skilled assistance and Dr.
Manfred Keller (both Universität Freiburg) for valuable spectrosco-
pic support.
The a-branched trifluoroethyl dienoate 31 stemmed from
an HWE reaction in which we combined the lithium de-
rivative of the a-branched trifluoroethyl phosphonylace-
tate 41 with enynal 43 (Scheme 3).26 This reaction
furnished 31 with 73% yield as a 2E,4E:2Z,4E:2E,4Z
(90:6.4:3.6) mixture. This corresponds to 93.6% E selec-
tivity with respect to the newly established C2=C3 bond
and 3.6% loss of trans configuration at the previously
present C4=C5 bond. HWE reagent 41, which was not re-
ported before, was obtained in 44% yield27 from an
Arbusov reaction between P(OMe)3 and trifluoroethyl
bromopropionate 40.28
References and Notes
(1) Reviews: (a) Lohray, B. B. Tetrahedron: Asymmetry 1992,
3, 1317. (b) Johnson, R. A.; Sharpless, K. B. Asymmetric
Catalysis in Organic Synthesis; Ojima, I., Ed.; VCH: New
York, 1993, 227. (c) Kolb, H. C.; VanNieuwenhze, M. S.;
Sharpless, K. B. Chem. Rev. 1994, 94, 2483. (d) Poli, G.;
Scolastico, C. Methoden der Organischen Chemie (Houben-
Weyl), Vol. E21e, 4th ed.; Helmchen, G.; Hoffmann, R. W.;
Synlett 2006, No. 16, 2641–2645 © Thieme Stuttgart · New York

2644
J. Schmidt-Leithoff, R. Brückner
LETTER
Mulzer, J.; Schaumann, E., Eds.; Thieme: Stuttgart, 1995,
4547. (e) Johnson, R. A.; Sharpless, K. B. Catalytic
Asymmetric Synthesis, 2nd ed.; Ojima, I., Ed.; Wiley-VCH:
New York, 2000, 357–389. (f) Bolm, C.; Hildebrand, J. P.;
Muñiz, K. Catalytic Asymmetric Synthesis, 2nd ed.; Ojima,
I., Ed.; Wiley-VCH: New York, 2000, 399–428.
(g) Zaitsev, A. B.; Adolfsson, H. Synthesis 2006, 1725.
Rationalizations of the absolute configuration of AD
products: (h) Empirically: Kolb, H. C.; Andersson, P. G.;
Sharpless, K. B. J. Am. Chem. Soc. 1994, 116, 1278. (i)
Calculationally: Moitessier, N.; Henry, C.; Len, C.;
Chapleur, Y. J. Org. Chem. 2002, 67, 7275.
34, 4975. (c) Caddick, S.; Shanmugathasan, S.; Brasseur,
D.; Delisser, V. M. Tetrahedron Lett. 1997, 38, 5735.
(d) Liu, B.; Chen, M. J.; Lo, C.-Y.; Liu, R.-S. Tetrahedron
Lett. 2001, 42, 2533. (e) Gardiner, J. M.; Giles, P. E.;
Martin, M. L. M. Tetrahedron Lett. 2002, 43, 5415.
(15) An ancillary observation of the same effect was our failure
to dihydroxylate trans-1,6-bis(trimethylsilyl)-3-hexene-1,5-
diyne with K2OsO2(OH)4–NMO, K2OsO2(OH)4–
(DHQ)2PHAL–K3Fe(CN)6 or KMnO4: Schmidt-Leithoff J.;
Ph.D. Dissertation; Universität Freiburg, 2006.
(16) We observed 5% more g,d- and 5% less a,b-dihydroxylation
in the mono-ADs of dienoates 1 and 3 than in the hands of
Sharpless et al.3 (see Scheme 1). This might be due to our
catalyst/ligand ratio being 1:5 and Sharpless’ being 1:1 and
1:2, respectively.
(2) (a) (6Z,9Z,11E)-6,9,11-Henicosatriene: Fernandez, R. A.;
Kumar, P. Tetrahedron 2002, 58, 6685. (b) (6E,8E)-6,8-
Tetradecadiene: Arizza, X.; Fernández, N.; Garcia, M.;
López, M.; Montserrat, L.; Ortiz, J. Synthesis 2004, 128.
(c) (8E,10E)-8,10-Dodecadienyl acetate and (2E,4E)-2,4-
hexadienyl benzoate: ref. 3. (d) (2E,4E)-2,4-Hexadiene,
(2E,4E)-2,dimethyl-2,4-hexadiene, and (2E,4Z)-2,4-
hexadiene: ref. 4.
(3) Becker, H.; Soler, M. A.; Sharpless, K. B. Tetrahedron
1995, 51, 1345.
(4) Xu, D.; Crispino, G. A.; Sharpless, K. B. J. Am. Chem. Soc.
1992, 114, 7570.
(17) 1,1,1,3,3,3-Hexafluoroisopropyl (2E,4E)-2-Methyl-7-
phenyl-2,4-heptadien-6-ynoate (33): 1H NMR (400.1
MHz, CDCl3, TMS; 4% 2Z,4E-isomer): d = 2.07 (dd,
4J2-Me,3 = 1.4 Hz, 5J2-Me,4 = 0.5 Hz, 2-CH3), 5.88 (sept, J1¢¢,F
=
6.2 Hz, 1¢¢-H), 6.28 (br d, J5,4 = 15.3 Hz, 5-H), 6.97 (dd,
J4,5 = 15.4 Hz, J4,3 = 11.7 Hz, 4-H), 7.32–7.38 (m, 3¢-H,
4¢-H, 5¢-H), partly superimposed by 7.38 (ddq, J3,4 = 11.7
Hz, 4J3,5 = 1.4 Hz, 4J3,2-Me = 0.9 Hz, 3-H), 7.43–7.52 (m, 2¢-
H, 6¢-H). HRMS (EI, 70 eV): m/z [M]+ calcd for C17H12F6O2:
362.0741; found: 362.0735.
(5) Zhang, Y.; O’Doherty, G. A. Tetrahedron 2005, 61, 6337.
(6) Schmidt-Leithoff, J.; Brückner, R. Helv. Chim. Acta 2005,
88, 1943.
(18) 1,1,1,3,3,3-Hexafluoroisopropyl (E,4S,5S)-4,5-
Dihydroxy-2-methyl-7-phenyl-2-hepten-6-ynoate (34):
K3Fe(CN)6 (273 mg, 828 mmol, 3.0 equiv), (DHQ)2PHAL
(21.5 mg, 27.6 mmol, 10 mol%), K2CO3 (114 mg, 828 mmol,
3.0 equiv), and MeSO2NH2 (26.3 mg, 276 mmol, 1.0 equiv)
were suspended in t-BuOH–H2O (4 mL:5 mL) at 0 °C.
K2Os(OH)4O2 (5.1 mg, 13.8 mmol, 5 mol%) and a solution of
33 (100 mg, 276 mmol) in t-BuOMe (2 mL) were added to
the reaction mixture. After stirring at 0 °C for 2 d sat. aq
Na2S2O3 (10 mL) was added. The resulting mixture was
stirred at r.t. for 30 min, the organic phase separated and
extracted with EtOAc (4 × 15 mL). The combined organic
phases were dried with Na2SO4. After evaporation of the
solvent the residue was purified by flash chromatography7
(eluent: cyclohexane–EtOAc, 3:1) giving the title compound
[43.5 mg, 40% of an inseparable E:Z mixture (95.4:4.6)] as
a colorless oil. 1H NMR (400.1 MHz, CDCl3, TMS; 4.6%
2Z-isomer): d = 2.07 (d, 4J2-Me,3 = 1.5 Hz, 2-CH3), 2.66, 2.81
(2 × br s, 4-OH, 5-OH), 4.58 (d, J5,4 = 7.0 Hz, 5-H), 4.66
(incompletely resolved dd, J4,3 = 8.3 Hz, J4,5 = 7.0 Hz, 4-H),
5.85 (sept, J1¢¢,F = 6.1 Hz, 1¢¢-H), 6.94 (dq, J3,4 = 8.4 Hz,
4J3,2-Me = 1.4 Hz, 3-H), 7.29–7.40 (m, ArH). HRMS (EI, 70
eV, fragment 1): m/z [M – C9H6O]+ calcd for C8H8F6O3:
266.0377; found: 266.0373. HRMS (EI, 70 eV, fragment 2):
m/z [M – C8H7F6O3]+ calcd for C9H7O: 131.0497; found:
131.0495.
(7) Still, W. C.; Kahn, M.; Mitra, A. J. Org. Chem. 1978, 43,
2923.
(8) All new compounds gave satisfactory 1H NMR and 13
C
NMR spectra and provided correct combustion analyses (5–
10, 13–15, 17, 21, 22, 25, 27, 29, 30) or high-resolution mass
spectra (16, 18–20, 23, 24, 26, 28, iso-30, 31–35, iso-32, 41).
(9) (a) Bennani, Y. L.; Sharpless, K. B. Tetrahedron Lett. 1993,
34, 2079. (b) See also: Tholander, J.; Carreira, E. M. Helv.
Chim. Acta 2001, 84, 613.
(10) Sharpless, K. B.; Amberg, W.; Bennani, Y. L.; Crispino, G.
A.; Hartung, J.; Jeong, K.-S.; Kwong, H.-L.; Morikawa, K.;
Wang, Z.-M.; Xu, D.; Zhang, X.-L. J. Org. Chem. 1992, 57,
2768.
(11) It is interesting to note that while we mono(dihydroxylated)
methyl dienoate 5 asymmetrically (see text) we could not
realize its racemic dihydroxylation at identical substrate
concentration using K2Os(OH)4O2 (10 mol%) and
NMO·H2O12 (1.2 equiv) in t-BuOH–H2O (1:1) over the
course of 4 d. This seems to imply that the asymmetric
dihydroxylation of compound 5 benefits from a considerable
ligand accelerating effect by the added amine.13 It was only
for this reason that all cis,vic-dihydroxylations of our study
were undertaken as asymmetric dihydroxylations and their
ee values considered unimportant and thus undetermined
[except for compounds 14 (99% ee) and 34 (84% ee)]. All
asymmetric dihydroxylations of our study were performed
with the AD-mix a ligand, for example, with (DHQ)2PHAL,
and never with the AD-mix b ligand, for example, with
(DHQD)2PHAL. The reason is that Sharpless et al. (ref. 3)
found decreased g,d:a,b dihydroxylation ratios employing
AD-mix b instead of AD-mix a for the dihydroxylation of
a,b,g,d-unsaturated esters 1 (→ 2:iso-2 = 83:17 instead of
87:13) or 3 (→ 4:iso-4 = 56:44 instead of 60:40).
(19) The only exception was hexafluoroisopropyl ester 33.17 It
was obtained in 96% yield by a carbodiimide-mediated
esterification of hexafluoroisopropanol with the carboxylic
acid obtained from the saponification of ethyl ester 29.
(20) Trifluoroethyl bromoacetate(35) was obtained in 75% yield
by an H2SO4-catalyzed esterification from bromoacetic acid
and trifluoroethanol (2.0 equiv). Previously, 35 was obtained
by trifluoroethanolysis of ethyl bromoacetyl chloride in 81%
yield.21
(12) VanRheenen, V.; Kelly, R. C.; Cha, D. F. Tetrahedron Lett.
1976, 1973.
(21) Morphy, J. R.; Rankovic, Z.; York, M. Tetrahedron 2003,
59, 2137.
(13) Schröder, M. Chem. Rev. 1980, 80, 187.
(22) Trifluoroethyl(dimethylphosphonyl)acetate(38) has not
been previously described and was prepared by the Arbusov
reaction (Scheme 2).20 This reaction is higher yielding (98%)
than the synthesis of the analogous trifluoroethyl(diethyl-
phosphonyl)acetate by treatment of (diethylphos-phono)
(14) Compare the diminished yields of ADs of alkynyl- versus
alkyl-substituted alkenes: (a) Jeong, K. S.; Sjö, P.;
Sharpless, K. B. Tetrahedron Lett. 1992, 33, 3833. (b)Tani,
K.; Sato, Y.; Okamoto, S.; Sato, F. Tetrahedron Lett. 1993,
Synlett 2006, No. 16, 2641–2645 © Thieme Stuttgart · New York

LETTER
Regioselective cis,vic-Dihydroxylation of a,b,g,d-Unsaturated Carboxylic Esters
2645
acetic acid first with SOCl2 and then with trifluoro-ethanol
(S = 58%): Birnbaum, J. C.; Busche, B.; Lin, Y.; Shaw, W.
J.; Fryxell, G. E. Chem. Commun. 2002, 1374.
6.22 (ddd, J5,4 = 15.4 Hz, 4J5,3 = 6J5,2¢/6¢ = 0.7 Hz, 5-H), 6.97
(J4,5 = 15.4 Hz, J4,3 = 11.8 Hz, 4-H), 7.31 [dqd, in part
superimposed by m (3¢-H, 4¢-H, 5¢-H), J3,4 ª 11.8 Hz,
4J3,2-Me = 1.4 Hz, 4J3,5 = 0.9 Hz, 3-H], 7.32–7.37 (m, 3¢-H,
4¢-H, 5¢-H), 7.44–7.51 (m, 2¢-H, 6¢-H). HRMS (EI, 70 eV):
m/z [M]+ calcd for C16H13F3O2: 294.0868; found: 294.0867.
(27) 2,2,2-Trifluoroethyl 2-(Dimethoxyphosphonyl)prop-
ionate (41): Neat 2,2,2-trifluoroethyl 2-bromopropionate
(4.58 g, 19.5 mmol) was heated at 60 °C while trimethyl
phosphite (2.99 mL, 3.14 g, 25.3 mmol, 1.3 equiv) was
added slowly. The resulting solution was then heated at 180
°C for 4 h. Distillation (bp 25–30 °C/0.45 mbar) afforded the
title compound (2.27 g, 44%). 1H NMR (400.1 MHz, CDCl3,
TMS): d = 1.48 (dd, J3,P = 17.7 Hz, J3,2 = 7.3 Hz, 3-H3), 3.15
(dq, 2J2,P = 23.8 Hz, J2,3 = 7.3 Hz, 2-H), 3.79, 3.80 (2 × d,
JOMe,P = 11.0 Hz, 2 × OCH3), AB signal (dA = 4.49, dB = 4.55,
(23) Zhu, X.-F.; Henry, C. E.; Wang, J.; Dudding, T.; Kwon, O.
Org. Lett. 2005, 7, 1387.
(24) The phosphonium bromide precursor of ylide 36 was
prepared from PPh3 and 2,2,2-trifluoroethyl bromoacetate
(35)20 in two steps and 100% overall yield. The same ylide
was similarly obtained by Kwon et al.23 but used en route to
allenic carboxylic esters and not in a Wittig reaction.
(25) Trifluoroethyl dienoate 13 was obtained as a mixture of
2E,4E-13 and 2E,4Z-13 isomers (98.5:1.5), which was
inseparable by flash chromatography on silica gel.7 The
formation of 2E,4Z-13 can be explained by an isomerization
of the C3=C4 bond.
(26) 2,2,2-Trifluoroethyl (2E,4E)-2-Methyl-7-phenyl-2,4-
heptadien-6-ynoate (31): At –78 °C n-BuLi (2.5 M in
hexane, 1.23 mL, 3.07 mmol, 1.4 equiv) was added to a
solution of 41 (694 mg, 2.63 mmol, 1.2 equiv) in THF (15
mL). After 10 min a solution of 43 (342 mg, 2.19 mmol) in
THF (10 mL) was added. Stirring was continued at –78 °C
for 30 min and at 0 °C for another 2 h. Quenching by adding
aq NH4Cl (10 mL), phase separation, extraction of the aq
phase with Et2O (3 × 15 mL), drying of the combined
organic phases with Na2SO4, and purification of the crude
product by flash chromatography7 (eluent: cyclohexane–
EtOAc, 5:1) furnished the title compound (73%) as a
2E,4E:2Z,4E:2E,4Z (90:6.4:3.6) mixture. 1H NMR (400.1
MHz, CDCl3, TMS): d = 2.04 (dd, 4J2-Me,3 = 1.4 Hz,
5J2-Me,4 = 0.5 Hz, 2-CH3), 4.56 (q, J1¢¢,F = 8.5 Hz, 1¢¢-H2),
JAB = 12.7 Hz, A and B peaks in addition split to q by J1¢,F =
8.3 Hz, B peaks further split to d by unassigned J = 0.5 Hz,
1¢-H2). HRMS (EI, 70 eV, fragment 1): m/z [M – OC3H3]+
calcd for C4H9F3O4P: 209.0190; found: 209.0187. HRMS
(EI, 70 eV, fragment 2): m/z [M – CH2CF3]+ calcd for
C5H10O5P: 181.0266; found: 181.0263. HRMS (EI, 70 eV,
fragment 3): m/z [M – OCH2CF3]+ calcd for C5H10O4P:
165.0317; found: 165.0316.
(28) Compound 40 was obtained in 78% yield by esterification of
2-bromopropionic acid. A two-step synthesis of 40 via acid
chloride formation from 2-bromopropionic acid followed by
trifluoroethanolysis yielded 83% of 40: Aggarwal, V. K.;
Jones, D. E.; Martin-Castro, A. M. Eur. J. Org. Chem. 2000,
2939.
Synlett 2006, No. 16, 2641–2645 © Thieme Stuttgart · New York