Catalytic enantioselective homoaldol reactions using binol titanium(IV) fluoride catalysts

Article abstract of DOI:10.1055/s-2003-37122

Titanium (IV) fluoride catalysts, prepared by the combination of (R)-2,2′-binaphthol and TiF₄, are effective for promoting the homoaldol addition of 1-ethoxy-1-(trimethylsilyloxy)cyclopropane to aldehydes. The reactions proceed with ee's of up to 72% and are effective with a range of aldehyde substrates. The reactions show greatly improved enantioselectivity when compared to those catalyzed by titanium(IV) triflates. The increase in selectivity is presumed to result from the elimination of deleterious silicon cocatalysis.

Full text of DOI:10.1055/s-2003-37122

390
LETTER
Catalytic Enantioselective Homoaldol Reactions Using Binol Titanium(IV)
Fluoride Catalysts
Catalytic
Enant
.
H
D
omoaldol Reaction
i
s
ane Burke, Ngiap Kie Lim, James L. Gleason*
Department of Chemistry, McGill University, 801 Sherbrooke St. West, Montreal, Quebec, H3A 2K6, Canada
Fax jim.gleason@mcgill.ca
Received 20 November 2002
towards homoenolate addition. Although this co-catalysis
Abstract: Titanium (IV) fluoride catalysts, prepared by the combi-
nation of (R)-2,2′-binaphthol and TiF4, are effective for promoting
the homoaldol addition of 1-ethoxy-1-(trimethylsilyloxy)cyclo-
propane to aldehydes. The reactions proceed with ee’s of up to 72%
and are effective with a range of aldehyde substrates. The reactions
show greatly improved enantioselectivity when compared to those
catalyzed by titanium(IV) triflates. The increase in selectivity is
presumed to result from the elimination of deleterious silicon co-
catalysis.
accelerated the reaction, it prevented the formation of a
closed transition state, which is presumably a necessary
requirement for high enantioselectivity.
In order to prevent the release of a highly Lewis acidic
species during the catalytic process, we have explored the
use of titanium(IV) fluorides as catalysts in this reaction.
It was expected that any trimethylsilyl cation generated in
this reaction would be trapped as the relatively non-acidic
trimethylsilyl fluoride, thus precluding silicon co-cataly-
sis. Alkoxytitanium(IV) fluorides have been developed by
Carreira for the catalysis of allylsilane and trimethylalu-
minum additions to aldehydes.4 The catalyst was prepared
by combining (R)-2,2′-binapthol (20 mol%) and TiF4 (10
mol%) in acetonitrile. After evaporation, the resulting
red-brown solid was suspended in a mixture of CDCl3–
CD3CN (3:1) and treated with a proton scavenger. Either
allyltrimethylsilane or cyclopropane 1 may be used as the
proton scavenger, with the former generally affording a
more active catalyst. The homoaldol reaction was con-
ducted by combining cyclopropane 1 and benzaldehyde
with the catalyst in CDCl3–CD3CN (3:1) at room temper-
ature. The reaction was monitored by 1H NMR and over a
period of 18 h, the silylated homoaldol adduct 3 was ob-
served to form (Scheme 1). After work-up, the crude
product was treated with p-TsOH in benzene to provide
lactone 4. Analysis of the product by chiral capillary GC
analysis (Chirasil-dex column) indicated that the product
had formed with 48% enantiomeric excess. Repeating the
reaction in a series of solvents (Table 1) revealed that pure
acetonitrile afforded the highest enantioselectivity (69%
ee).5,6
Key words: homoenolates, homoaldol additions, asymmetric ca-
talysis, titanium(IV) fluorides
In the field of asymmetric catalysis, the aldol reaction fea-
tures prominently because of its ability to simultaneously
form a carbon-carbon bond and to set up to two new
stereocenters. Catalytic asymmetric homoaldol reactions,
in contrast, are virtually unknown. This reflects the great-
er difficulty both in forming homoenolates and in success-
fully executing their subsequent addition to aldehydes and
ketones.1 To date, the highest enantioselectivity attained
in a catalytic enantioselective homoaldol reaction was
30% ee using a pre-formed titanium homoenolate and
Seebach’s Taddol-Ti(i-PrO)2 catalyst.2 In this letter, we
report a greatly improved catalytic enantioselective ho-
moaldol reaction based on titanium(IV) fluoride catalysts.
We previously reported a catalytic homoaldol reaction
which employed titanium(IV) triflates as catalysts.3 The
advantage of this catalytic system is that it did not require
the prior formation of a metal homoenolate. The desired
reactive intermediate was instead formed in situ via ring-
opening of commercially available 1-ethoxy-1-(trimeth-
ylsilyloxy)cyclopropane. Although the reaction was effi-
cient, only low enantioselectivities were observed using
(R)-2,2′-binaphthol as a ligand. The low enantioselectivi-
ty in this reaction was attributed to the formation of tri-
methylsilyl triflate, which served to activate the aldehyde
Examining the reaction further, it was found that the high-
est enantioselectivities were observed at room tempera-
ture, with the use of either higher or lower temperatures
leading to markedly lower selectivity (Table 2, entries
O
OTMS
pTsOH
Ti catalyst
+ PhCHO
(10 mol%)
EtO OTMS
O
EtO2C
Ph
PhH
Ph
4
1
2
3
Scheme 1
Synlett 2003, No. 3, Print: 19 02 2003.
Art Id.1437-2096,E;2002,0,02,0390,0392,ftx,en;S03702ST.pdf.
© Georg Thieme Verlag Stuttgart · New York
ISSN 0936-5214

LETTER
Catalytic Enantioselective Homoaldol Reactions
391
Table 3 Homoaldol Reactions of Aromatic and Acetylenic Alde-
Table 1 Effect of Solvent on the Stereoselectivity of the Homoaldol
Reaction (Equation 1).
hydesa
O
Entry
Solventa
3:1 CDCl3/CD3CN
CDCl3
eeb (%)
1. 10 mol% Ti catalyst
2. TsOH, benzene
EtO OTMS
O
+ RCHO
1
2
3
4
5
6
7
8
9
48
0
R
Entry
R
eeb (%)
65
yield (%)
80
CH3CN
PhCN
69
35
17
5
1
2
3
4
5
6
7
8
9
C6H5-
1-C10H9-
2-C10H9-d
4-ClC6H4-
4-BrC6H4-
4-CF3C6H4-
4-CH3C6H4-
4-CH3OC6H4-
C6H5C≡C-
55
59c
85
Et2O
61e
45
THF
62
C6H6
0
43
69
CH3NO2
DMF
7
40
69
8
a All reactions were carried out using 10 mol% catalyst and 1.0 equiv
silyloxycyclopropane in the indicated solvent.
18
59
0
ND
41
b Determined by chiral GC analysis (Chirasil-dex column).
61e
1–4). The yield of the reaction using a one to one ratio of
1 and benzaldehyde was only 36%, but this could be in-
creased by employing a modest excess of cyclopropane 1
(entries 5 and 6) and by running the reaction at slightly
higher concentrations. The optimum conditions utilized
1.5 equivalents of 1 and a reaction concentration of ap-
proximately 0.7 M.
a Reactions were allowed to proceed for 7 d under conditions cited in
Table 2, entry 5 for a period of 168 h.
b Determined by chiral GC analysis (Chirasil-Dex column).
c Based on recovered starting material.
d The catalyst was aged for 2.5 h for this example.
e Determined by Chiral HPLC analysis [Pirkle covalent (S,S) Whelk
–0 1].
um(IV) triflate catalysts,3 aliphatic and a,b-unsaturated
aldehydes were not useful substrates for the reaction.
Table 2 Optimization of Temperature and Concentration
(Scheme 1).a
Entry
T
(°C)
1:PhCHO
[PhCHO]
(M)
eeb
(%)
Yield
(%)
Details regarding the catalytic cycle of this process remain
to be elucidated. Although the intermediacy of a metal ho-
moenolate seems likely, as with the titanium(IV) triflate
catalyzed reaction, it was not possible to observe ho-
1
2
3
4
5c
6
–30
0
1: 1
0.55
0.55
0.55
0.55
0.52
0.69
0
31
66
57
72
65
ND
ND
36
1
1: 1
moenolate formation by H NMR. The products are in-
variably isolated as the trimethylsilyl ethers. Given the
low probability of silylation by trimethylsilyl fluoride,
there seems to be two likely mechanisms for homoaldola-
te silylation. The first would be direct silicon transfer dur-
ing the ring opening of 1 by a product homoenolate
titanium complex, simultaneously forming 3 and a new
homoenolate. Alternatively, the binaphthol ligand might
play a part as a silicon shuttle. In this regard, it is impor-
tant to note that the optimum reaction conditions require a
2:1 ratio of 2,2′-binaphthol to TiF4. The use of a 1:1 ratio
results not only in lower yields and ee’s, but also results in
formation of a significant amount of lactone 4, indicating
that silicon transfer is a problem under these conditions.
25
50
25
25
1: 1
1: 1
ND
66d
80d
1.5: 1
1.5: 1
a Reactions were conducted over a 24 h period with a catalyst aging
time of 1 h prior to addition of 1 and PhCHO, except as noted.
b Determined by Chiral GC analysis (Chirasil-Dex column).
c The catalyst was aged for 2.5 hours rather than one hour prior to ad-
dition of 1 and PhCHO.
d Reaction time 3 d.
A series of aromatic and acetylenic aldehydes were sur-
veyed under the optimized reaction conditions.7 As can be
seen from the results in Table 3, good yields and modest
selectivities were observed with simple aromatic alde-
hydes and those containing electron withdrawing groups
at the 4-position. In contrast, electron donating groups at
the 4-position (entries 7 and 8) significantly reduce the
enantioselectivity of the reaction. As with the titani-
In conclusion, we have developed a catalytic asymmetric
homoaldol reaction, which utilizes an alkoxytitanium(IV)
fluoride catalyst. The enantioselectivities, while modest,
are the highest reported for this class of reactions. The
method works well for a series of aromatic and acetylenic
aldehydes and gives the highest enantioselectivities for
electron-neutral and electron-poor aromatic aldehydes.
Synlett 2003, No. 3, 390–392 ISSN 0936-5214 © Thieme Stuttgart · New York

392
E. D. Burke et al.
LETTER
binaphthol (56 mg, 0.195 mmol, 0.2 equiv) in acetonitrile
(1.5 mL) in a flame dried Schlenk flask. The resulting red
mixture was stirred at 21 °C for 1 h and then the solvent was
removed in vacuo. The resulting oil was redissolved in
acetonitrile (1.5 mL), allyltrimethylsilane (62 µL, 0.392
mmol, 0.4 equiv) was added and the resulting solution was
allowed to stir for 3 h during which time a precipitate was
observed to form. To this mixture was added 1-ethoxy-1-
(trimethylsilyloxy)cyclopropane (300 µL, 1.50 mmol, 1.5
equiv), and after 10 min, benzaldehyde (100 µL, 0.99 mmol,
1 equiv) was added. This final solution was continuously
stirred over 4 days. The reaction was quenched by addition
of 1 M HCl (30 mL) and the products were extracted into
ethyl acetate (2 × 30 mL). The organic layer was washed
once with brine (30 mL), dried over Na2SO4, filtered and
concentrated. The crude reaction mixture was dissolved in
benzene (6 mL) and p-TsOH(cat) was added. The reaction
mixture was stirred overnight. The reaction was quenched by
addition of saturated NaHCO3 (30 mL) and the product was
extracted into ethyl acetate (2 × 30 mL). The organic layer
was washed once with brine (30 mL), dried over Na2SO4,
filtered and concentrated. Purification of the residue by
column chromatography on silica gel using 30 % hexanes in
methylene chloride as eluent afforded 128 mg of the lactone
(66% yield, 72% ee). 1H NMR (CDCl3) δ 7.24–7.39 (m, 5
H), 5.46 (dd, 1 H, J = 7.9 Hz, 6.2 Hz), 2.54–2.67 (m, 3 H),
2.04–2.22 (m, 1 H); 13C NMR (CDCl3) δ 176.8, 139.2,
128.5, 125.1, 81.0, 30.7, 28.7. Anal. Calcd for C10H10O2: C,
74.06; H, 6.21. Found: C, 73.73; H, 6.02.
Acknowledgement
We thank NSERC and Boehringer Ingelheim Inc. for funding of this
research. E. D. B thanks NSERC for a postgraduate fellowship.
References
(1) For reviews on the homoaldol reaction, see: (a) Crimmins,
M. T.; Nantermet, P. G. Org. Prep. Proc. 1993, 25, 43.
(b) Kuwajima, I.; Nakamura, E. In Comprehensive Organic
Synthesis, Vol 2; Trost, B. M.; Fleming, I., Eds.;
Pergammon: Oxford., 1991, 441.
(2) Weber, B.; Seebach, D. Tetrahedron 1994, 50, 7473.
(3) Martins, E. O.; Gleason, J. L. Org. Lett. 1999, 1, 1643.
(4) (a) Gauthier, D. R. Jr.; Carreira, E. M. Angew. Chem., Int.
Ed. Engl. 1996, 35, 2363. (b)Bode, J. W.; Gauthier, D. R.
Jr.; Carreira, E. M. Chem. Commun. 2001, 24, 2560.
(b) Pagenkopf, B. L.; Carreira, E. M. Tetrahedron Lett.
1998, 39, 9593.
(5) The absolute stereochemistry was established as R by
comparison of the optical rotation of 4 with literature values:
Gutman, A. L.; Zuobi, K.; Bravdo, T. J. Org. Chem. 1990,
55, 3546.
(6) A survey of other ligands, including 3,3′-disubstituted
binaphthols, F8-Binol and TADDOL did not afford any
increase in enantioselectivity.
(7) Sample Experimental Procedure. A solution of
titanium(IV) fluoride (12 mg, 0.095 mmol, 0.1 equiv) in
acetonitrile (1.5 mL) was added to a solution of (R)-2,2′-
Synlett 2003, No. 3, 390–392 ISSN 0936-5214 © Thieme Stuttgart · New York