Room temperature asymmetric allylic trifluoromethylation of Morita-Baylis-Hillman carbonates

Article abstract of DOI:10.1021/ol202572u

(DHQD)₂PHAL-catalyzed asymmetric allylic trifluoromethylation of Morita-Baylis-Hillman adducts using a Rupert-Prakash reagent is reported. This transformation provided the S_N2′ trifluoromethylated products with good yields and excell

Full text of DOI:10.1021/ol202572u

ORGANIC
LETTERS
2011
Vol. 13, No. 22
6082–6085
Room Temperature Asymmetric Allylic
Trifluoromethylation of MoritaꢀBaylisꢀ
Hillman Carbonates
Yun Li,† Fang Liang,† Qian Li,† Yao-chang Xu,† Quan-Rui Wang,‡ and Lei Jiang*,†
China Novartis Institutes for BioMedical Research Co. Ltd., Building 8,
lane 898 Halei road, Shanghai 201203, China, and Department of Chemistry,
Fudan University, 220 Handan Road, Shanghai 200433, China
lei-1.jiang@novartis.com
Received September 23, 2011
ABSTRACT
(DHQD)2PHAL-catalyzed asymmetric allylic trifluoromethylation of MoritaꢀBaylisꢀHillman adducts using a RupertꢀPrakash reagent is reported.
This transformation provided the SN20 trifluoromethylated products with good yields and excellent enantioselectivities at room temperature. It was
also found that the reaction could be accelerated using acetonitrile as cosolvent.
Trifluoromethylated compounds have gained growing
interest owing to their unique physical and chemical
properties such as great stability, high electronegativity,
and good lipophilicity. These properties may have con-
tributed to the enhanced bioavailability and metabolic
stability found in fluorinated pharmaceutical agents,1
and therefore increasing efforts have been devoted in this
particular field.2 To date, significant progress has been
made in the arena of nucleophilic 1,2-trifluoro-
methylation3 and R-trifluoromethylation4 of carbonyl
compounds including their asymmetric version.5 How-
ever, the reported methods of introducing a trifluoro-
methyl group, especially their asymmetric versions, are
still at their infancy and cannot meet the growing needs
from both pharmaceutical and advanced material in-
dustry. To this end, there is a strong need to develop
more efficient and convenient trifluoromethylation
reactions.
Allylic alkylation is one of the most powerful ap-
proaches to construct carbonꢀcarbon and carbonꢀ
heteroatom bonds.6 Recently, the allylic alkylation of
MoritaꢀBaylisꢀHillman (MBH) adducts catalyzed by
organic Lewis base have become an attractive strategy
† China Novartis Institutes for BioMedical Research Co., Ltd.
‡ Fudan University.
(1) For recent reviews, see: (a) Yamazaki, T.; Taguchi, T.; Ojima, I.
In Fluorine in Medicinal Chemistry and Chemical Biology; Wiley-Black-
well: Chichester, 2009; p 3. (b) Hagmann, W. K. J. Med. Chem. 2008, 51,
4359. (c) Muller, K.; Faeh, C.; Diederich, F. Science 2007, 317, 1881. (d)
Biomedical Frontiers of Fluorine Chemistry; Ojima, I., McCarthy, J. R.,
Welch, J. T., Eds.; American Chemical Society: Washington, DC, 1996.
(e) Smart, B. E. Chem. Rev. 1996, 96, 1555. (f) Resnati, G.; Soloshonok,
V. A. Tetrahedron 1996, 52, 1.
(2) (a) Nie, J.; Guo, H. -C.; Cahard, D.; Ma, J.-A. Chem. Rev. 2011,
111, 455. (b) Furuya, T.; Kamlet, A. S.; Ritter, T. Nature 2011, 473, 470.
(c) Ma, J.-A.; Cahard, D. J. Fluorine Chem. 2007, 128, 975. (d) Tozer,
M. J.; Herpin, T. F. Tetrahedron 1996, 52, 8619.
to construct
a large variety of multifunctional
compounds.7 Enlightened by these findings, we postu-
lated that (Scheme 1) a Lewis base catalyzed allylic
trifluoromethylation could be achieved via a SN20 reac-
tion. The strong basicity of tert-butoxy anion may
(3) (a) Prakash, G. K. S.; Wang, Y.; Mogi, R.; Hu, J.; Mathew, T.;
Olah, G. A. Org. Lett. 2010, 12, 2932. (b) Singh, R. P.; Kirchmeier, R. L.;
Shreeve, J. N. M. Org. Lett. 1999, 1, 1047. (c) Russell, J.; Roques, N.
Tetrahedron 1998, 54, 13771. (d) Dilman, A. D.; Levin, V. V.; Belyakov,
P. A.; Struchkova, M. I.; Tartakovsky, V. A. Tetrahedron Lett. 2008, 49,
4352. (e) Chang, Y.; Cai, C. J. Fluorine Chem. 2005, 126, 937.
(4) (a) Mizuta, S.; Shibata, N.; Akiti, S.; Fujimoto, H.; Nakamura,
S.; Toru, T. Org. Lett. 2007, 9, 3707. (b) Mizuta, S.; Shibata, N.; Ogawa,
S.; Fujimoto, H.; Nakamura, S.; Toru, T. Chem. Commun. 2006, 2575.
(c) Boechat, N.; Bastos, M. M. Curr. Org. Synth. 2010, 7, 403.
(5) (a) Nagib, D. A.; Scott, M. E.; MacMan, D. W. C. J. Am. Chem.
Soc. 2009, 131, 10875. (b) Allen, A. E.; MacMillan, D. W. C. J. Am.
Chem. Soc. 2010, 132, 4986.
(6) (a) Trost, B. M.; Van Vranken, D. L. Chem. Rev. 1996, 96, 395.
(b) Ojima, I. Catalytic Asymmetric Synthesis; Wiley: New York, 2000.
r
10.1021/ol202572u
Published on Web 10/25/2011
2011 American Chemical Society

Table 1. Identifying Catalyst for the Reactiona
Scheme 1. Proposed Lewis Base Catalyzed Allylic Trifluo-
methylation of MBH Carbonates
entry
catalyst
DABCO
solvent
yieldc (%)
1
DMF
DME
DME
DME
DMF
DME
DME
0
85
0
2
DABCO
3
none
4
PPh3
0
5
PPh3
(η3-C3H5PdCl)2
Pd þ i-Pr PhOX
0
regenerate the hard nucleophilic CF3 anion in situ.
During the preparation of this manuscript, Shibata
and co-workers reported the first allylic trifluoro-
methylation of MBH adducts in high yields.8 The four
enantioselective examples they described require long
reaction time and elevated temperature (ca. 5 days,
60 °C) and excess of Ruppert reagent (5 equiv). The
desired products were isolated with modest yield.
Herein, we describe an improved reaction condition
for the cinchona alkaloid catalyzed asymmetric allylic
trifluoromethylation of MBH carbonates using
Ruppert reagent. This method allows the reaction to
proceed under room temperature with much improved
yield.
Initially, with the Ruppert reagent as the trifluoro-
methyl source, we examined the reaction using 4-nitro-
MBH carbonate (1a) as the prototypical substrate and
DABCO or Ph3P as the Lewis base to test our specu-
lated transformation in different solvents (Table 1). To
our delight, we found that DABCO smoothly cata-
lyzed the allylic trifluorometylation in DME with very
good chemoselectivity and yield. In contrast, no de-
sired product was found when using Ph3P as catalyst or
DMF as solvent. Notably, when palladium-catalyzed
allylic alkylation conditions were employed9 (entries 6
and 7, Table 1), no desired product was found in the
reaction mixture. Encouraged by this preliminary re-
sult, we envisaged that an appropriate chiral tertiary
amine might be able to transfer its chirality to the
newly generated allylic carbon bearing a CF3 group
based on the mechanism depicted in Scheme 1. A
literature survey led us to cinchona and its derivatives,
6
7b
0
0
a Unless otherwise noted, reactions were performed with 0.1 mmol of
1a, 0.25 mmol of TMSCF3, 0.01 mmol of KF, and 0.01 mmol of catalyst
in 0.5 mL of solvent. b (η3-C3H5PdCl)2, 0.005 mmol, and (R)-(ꢀ)-
2-[2-(diphenylphosphino)phenyl]-4-isopropyl-2-oxazoline (i-Pr PhOX),
0.01 mmol, were used. c Isolated yield.
especially (DHQD)2PHAL, which was first reported
by Sharpless et al. in asymmetric dihydroxylation
reactions.10 This chiral tertiary amine base is usually
treated as an enantiomeric equivalent of DABCO and has
been applied in various asymmetric organic trans-
formations.11 We found that under the same condition as
that for DABCO, (DHQD)2PHAL was a less efficient
catalyst. The reaction hardly proceeded to full conversion
in DME even after 120 h of stirring at room temperature.
However, to our delight, high enantiomeric excess was
observed in the trial reactions. Elevating the reaction tem-
perature to 50 °C showed slight acceleration of reaction
rate, but the ee dropped from 94% to 79% (entries 1 and 2,
Table 2). When this reaction was performed under higher
concentration (0.5 M), the reaction could proceed to full
conversion with good yield and excellent ee value (95%)
at room temperature. Reaction rate could be further
accelerated with the substrate concentration reaching
1 M; however, the ee value started to drop to 89%. Con-
sidering charge could build up for the key reaction
intermediates, we decided to further investigate solvent
effects for this transformation.
Solvent screening did show that acetonitrile has signifi-
cant acceleration effect on this transformation. The reac-
tion could proceed to full conversion within 8 h;
however, the ee value was less satisfying (entry 7,
Table 2). It is also interesting to note that in this solvent
system, when the reaction was performed under lower
concentration and catalyst loading, we were still able to
drive the reaction to completion within 24 h with some
(7) (a) Cho, C.-W.; Krische, M. J. Angew. Chem., Int. Ed. 2004, 43,
6689. (b) Cui, H.-L.; Feng, X.; Peng, J.; Lei, J.; Jiang, K.; Chen, Y.-C.
Angew. Chem., Int. Ed. 2009, 48, 5737. (c) Feng, X.; Yuan, Y.-Q.; Cui,
H.-L.; Jiang, K.; Chen, Y.-C. Org. Biomol. Chem. 2009, 7, 3660. (d)
Hong, L.; Sun, W.; Liu, C.; Zhao, D.; Wang, R. Chem. Commun. 2010,
46, 2856. (e) Jiang, Y.-Q.; Shi, Y.-L.; Shi, M. J. Am. Chem. Soc. 2008,
130, 7202. (f) Sun, W.; Hong, L.; Liu, C.; Wang, R. Org. Lett. 2010, 12,
3914. (g) van Steenis, D. J. V. C.; Marcelli, T.; Lutz, M.; Spek, A. L.; van
Maarseveen, J. H.; Hiemstra, H. Adv. Synth. Catal. 2007, 349, 281.
(h) Zhang, S.-J.; Cui, H.-L.; Jiang, K.; Li, R.; Ding, Z.-Y.; Chen, Y.-C.
Eur. J. Org. Chem. 2009, 5804.
(8) Furukawa, T.; Nishimine, T.; Tokunaga, E.; Hasegawa, K.;
Shiro, M.; Shibata, N. Org. Lett. 2011, 13, 3972.
(9) Helmchen, G.; Pfaltz, A. Acc. Chem. Res. 2000, 33, 336.
(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.
J. Org. Chem. 1992, 57, 2768.
(11) (a) Whitehead, D. C.; Yousefi, R.; Jaganathan, A.; Borhan, B.
J. Am. Chem. Soc. 2010, 132, 3298. (b) Tian, S.-K.; Hong, R.; Deng, L.
J. Am. Chem. Soc. 2003, 125, 9900. (c) Poisson, T.; Gembus, V.; Dalla,
V.; Oudeyer, S.; Levacher, V. J. Org. Chem. 2010, 75, 7704. (d) Ogawa,
S.; Shibata, N.; Inagaki, J.; Nakamura, S.; Toru, T.; Shiro, M. Angew.
Chem., Int. Ed. 2007, 46, 8666. (e) Kobbelgaard, S.; Brandes, S.;
Jorgensen, K. A. Chem.;Eur. J. 2008, 14, 1464. (f) Bui, T.; Borregan,
M.; Barbas, C. F. J. Org. Chem. 2009, 74, 8935. (g) Bella, M.; Jorgensen,
K. A. J. Am. Chem. Soc. 2004, 126, 5672.
Org. Lett., Vol. 13, No. 22, 2011
6083

Table 2. Optimization of Reaction Conditions
Table 3. Optimization of the DMEꢀMeCN Bisolvent Ratioa
entry
ratio
time (h)
yield (%)
ee (%)
1
10:1
5:1
2:1
1:1
4:1
60
48
48
36
56
35
50
62
85
83
88
92
88
84
93
2
3
cat. (mol
%)
concn
(M)
time
(h)
yieldc
(%)
eed
(%)
4
5b
entry
solvent
a Reaction was carried out using 1a (0.5 mmol), Me3SiCF3
(1.25 mmol), and (DHQD)2PHAL (0.05 mmol) in solvent (1.7 mL) at
room temperature unless otherwise noted. b (DHQD)2PHAL (0.075 mmol)
and solvent (1 mL) were used.
1
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.2
0.1
0.1
0.1
0.1
none
0.20
0.20
0.50
1.0
DME
120
65
106
72
72
72
8
41
40
84
86
53
39
72
48
66
55
85
86
87
43
51
0
94
79
95
89
91
90
74
90
87
93
72
65
78
76
78
NT
2a
3
DME
DME
4
DME
5
1.0
THF
6
1.0
toluene
CH3CN
DCE
different electronic nature on the aromatic ring, such
as nitro, methoxy, alkyl, alkene, and halogen, were
tolerated (entries 1ꢀ6 and 9, Table 4; 32ꢀ86% yield,
87ꢀ94% ee). Generally, electron-withdrawing groups
on the aromatic ring facilitated the reaction, whereas
electron-donating substituents retarded the reaction.
For substrates bearing electron-donating groups, the
reaction could be modulated by increasing the amount
of acetonitrile in the bisolvent system (entry 2, Table 4).
Common protecting groups such as Ac, Boc, and allyl
(entries 7 and 8; 47ꢀ80% yield, 58ꢀ94% ee) were also
compatible in this system. Heteroaromatic ring sys-
tems such as furan, substituted furan, thiophene, thia-
zole, and quinoline were also viable replacements for
the benzene ring in MBH substrates (entries 10ꢀ12, 14,
and 15, Table 4; 54ꢀ84% yield, 48ꢀ92% ee). As
expected, the electron-richer thiophene substrate pro-
vided product 2p with lower yield and ee value than the
product 2n from 2-methyl thiazole substrate even
under 1:1 DME/acetonitrile ratio . It should be noted
that the styryl compound (entry 13, Table 4) gave
exclusively allylic trifluoromethylated product 2m12
instead of a diene product, which might be generated
via homoallylic trifluoromethylation of the active io-
nic intermediate. MBH substrates with alkyl substitu-
tion have also been examined (Scheme 2). However,
when 1p was used as the substrate and DABCO as the
catalyst, the elimination process was predominant,
likely resulted in intermediate 4a. Dimerization of 4a
could further lead to the cyclized product 3p, which
was the solely isolated product with very high yield
(93%). When sterically demanding cyclohexyl sub-
strate 1q was employed under the same reaction con-
ditions, the elimination process was not observed, and
the desired product 2q was successfully isolated in 72%
yield. However, the enantioselective trifluoromethyla-
tion of 1q could not be achieved13 even under long
reaction time and higher temperature (5 days, 50 °C).
This reaction is still under investigation.
7
1.0
8
1.0
72
72
72
8
9
1.0
EtOAc
dixoane
CH3CN
CH3CN
CH3CN
PhCN
C5H11CN
CH3CN
10
11
12
13b
14
15
16
1.0
0.50
0.30
0.30
0.30
0.30
0.30
14
24
65
65
48
a Reaction was performed at 50 °C. b Reaction was performed at
0 °C. c Isolated yield. d Determined by chiral SFC analysis.
sacrifice of enantioselectivity (entries 11ꢀ13, Table 2).
Lowering the reaction temperature to 0 °C did not
improve the ee value. Other cyano containing solvents
such as valeronitrile and benzonitrile failed to show any
reaction rate acceleration or enantioselectivity benefit.
In order to ensure that the acceleration effect is not a
result of the weak nucleophilicity of the cyano group, a
blank control reaction was carried out in acetonitrile in
the absence of the catalyst, and no detectable product
was formed in the reaction. These data suggest that the
remarkable rate acceleration of acetonitrile is not due
to its weak nucleophilicity, which may have a compe-
titive role against (DHQD)2PHAL, but through the
stabilization of the charged transition states.
With the facts that etherate solvent can provide good
yield and enantioselectivity but a slow reaction,
whereas the acetonitrile can accelerate the reaction
turnover with some sacrifice in enantioselectivity, it
was only logical for us to investigate the bisolvent
system. Thus the reactions were carried out in the
DME/acetonitrile hybrid solvent system with a varia-
tion of solvent ratios (Table 3). To our delight, with
1a as the substrate, an optimal DME/acetonitrile ratio
of 4:1 was identified. Under this condition, the reac-
tion could achieve 83% yield and 93% ee at room
temperature for 56 h when 15 mol % (DHQD)2PHAL
was used.
With this optimized reaction condition in hand, the
scope of this trifluoromethylation protocol was investi-
gated with a variety of MBH carbonates, and the
results are summarized in Table 4. Substituents with
(12) Structrue was assigned by COSY and HMBC. See Supporting
Information.
(13) Furukawa, T.; Kawazoe, J.; Zhang, W.; Nishimine, T.;
Tokunaga, E.; Matsumoto, T.; Shiro, M.; Shibata, N. Angew. Chem.,
Int. Ed. 2011, 123, 9858.
6084
Org. Lett., Vol. 13, No. 22, 2011

Table 4. Enatioselective Allylic Trimetylfluoromethylation of MBH Carbonates
a Sterochemistry was assigned by the chemical analogy of ref 8. b Isolated yield. c Enatiomeric excess was determined by chiral HPLC or SFC analysis.
d Using DME/MeCN (1:1) as solvent. e Reaction was also performed in DME/MeCN (1:1) conditions; 80% yield and 67% ee were obtained.
available Ruppert reagent and a bis-cinchona alkaloid
as catalyst has achieved good yield and excellent ee.
The kinetic of this reaction was largely dependent on
Scheme 2. Reaction Using Alkyl-Substitued Substratesa
the electronic nature of the aromatic ring in the MBH
substrates, with more electron-withdrawing groups
providing facile reactions. Acetonitrile was also found
to be able to accelerate the reaction rate, which allowed
an efficient enantioselective synthesis of this class of
compounds at room temperature.
Acknowledgment. We thank Mr. Jiangwei Zhang of
China Novartis for the MS and HRMS analytical support.
We also thank Dr. En Li of China Novartis for his support
and encouragement. Y.L. is grateful for the financial
support through the postdoctoral program from the Edu-
cational Office of Novartis Institutes for BioMedical
Research.
a Determined by 1H NMR spectroscopy.
Supporting Information Available. Experimental pro-
cedures and compound characterization data including
In summary, enantioselective allylic trifluoromethylation
of MoritaꢀBaylisꢀHillman adducts using commercially
HPLC (SFC) data. This material is available free of
charge via the Internet at http://pubs.acs.org.
Org. Lett., Vol. 13, No. 22, 2011
6085