3138
J. Li et al. / Tetrahedron Letters 55 (2014) 3137–3139
Table 1
Table 2
Impact of TiCl(OiPr)3 to CHCl3/ketone addition reactionsa
Chlorotitanium (IV) triisopropoxide promoted addition of chloroform to ketonesa
Product 2
Assay yieldb
no additive
(%)
Isolated yieldc Isolated yieldc
Entry
Ketones
Products
Isolated
with MgCl2
(%)
with TiCl(OiPr)3
(%)
yieldb (%)
HO
CCl3
HO
CCl3
1
2
3
1c
95
93
91
0
—
91
78
2c
N
Bn2N
HO
N
2a
HO CCl3
1d
CCl3
28
41
2d
2b
HO
CCl3
a
Reaction conditions: CHCl3 (5 equiv), n-BuLi (5 equiv), TiCl(OiPr)3 or MgCl2
1e
(0 equiv or 2 equiv), THF (0.5 M).
2e
b
Assay yield was determined by HPLC comparison with a standard solution of
HO
HO
purified product.
c
CCl3
Purified by flash chromatography on silica gel.
4
5
1f
84
93
2f
CCl3
HO
R1
O
TiCl(Oi-Pr)3, CHCl3, THF
then n-BuLi, -60 °C
CCl3
R2
1g
R1
R2
2g
2
HO CCl3
1
6
7
8
1h
1i
88
96
71
Scheme 2. Titanium promoted synthesis of trihalomethylcarbinol.
2h
HO CCl3
DBU, CHBr3/LiOH, CHBr3/LiHMDS, and CHCl3/n-BuLi were investi-
gated unsuccessfully. Most of the starting ketone remained unre-
acted under these conditions suggesting the reaction suffered
from poor reactivity and stability of the chloroform carbanion as
well as ketone enolization, which was apparent in the case of
2i
HO CCl3
1j
2j
CHBr3/LiHMDS, where trace amounts of
observed as a byproduct of the reaction.
a-bromination were
HO CCl3
Scouting for an experimentally simple solution to suppress
the enolization problem that would be amenable to the gener-
ation of multiple and diverse analogs, we turned our attention
to the use of organotitanium reagents. Organotitanium reagents
have been shown to be superior to traditional Grignard reagents
for additions to sterically hindered and/or enolizable ketones.14
Treatment of chloroform carbanion with TiCl(OiPr)3 at a low
temperature (À60 °C) followed by addition of the ketone start-
ing material 1a led to a clean conversion to the desired triha-
lomethyl-carbinol 2a. Looking for the simplest possible
experimental procedure, we explored the order of addition of
the reagents and found that simply combining the ketone,
CHCl3, and TiCl(OiPr)3 in cold THF and adding n-BuLi slowly
provided a similar result. The drastic improvement on substrate
1a encouraged us to explore the scope of this transformation.
We started by carrying out the same comparison of reaction
conditions on indanone 1b. By comparison, the reaction without
any additive only gave 28% assay yield, the reaction with
2 equiv MgCl2 gave 41% isolated yield, while the reaction with
2 equiv TiCl(OiPr)3 provided 83% assay yield and 78% isolated
yield (Table 1).
9
1k
1l
83c
45
2k
OH
CCl3
10
2l
a
Reaction conditions: CHCl3(5 equiv), n-BuLi (5 equiv), TiCl(OiPr)3 (2 equiv), THF
(0.5 M).
b
Purified by flash chromatography on silica gel.
Relative stereochemistry determined by comparision of 1H NMR with literature
c
data9b (dr > 9:1 by 1H NMR).
Cl
Cl
HO
CCl3
1) H2SO4, THF
NaOMe
Bn2N
N
Bn2N
N
2a
3a
OMe
2) NaOMe, MeOH,
then HCl
MeO
OMe
CO2Me
H+
The success of these experiments prompted us to examine the
improved reaction conditions on a variety of other enolizable
ketones (Scheme 2). The results summarized in Table 2 indicate
that the titanium-promoted addition reaction works well on a
number of ketones, including alkyl-aryl (entries 1–6), bisalkyl
Bn2N
N
4a
Bn2N
N
5a
Scheme 3. Dehydration and hydrolysis of trihalomethylcarbinol. Reagents and
conditions: (1) H2SO4, THF, 70 °C; (2) NaOMe, MeOH, 60 °C, then aq HCl, 80% for two
steps.
(entries 8–9), and
a,b-unsaturated ketones (entry 7), to give the
expected products in satisfactory yields. The presence of more ste-
rically demanding branched alkyl groups (entries 4, 5, and 9) had
no effect on the product formation as high yields were achieved
in all of these examples. This protocol was also applied to a highly
enolizable substrate, 2-indanone, to give desired trichloromethyl-
carbinol in moderate yield (entry 10).