Nucleophilic displacements of non-racemic α-trifluoromethyl benzylic triflates

Article abstract of DOI:10.1016/j.tet.2008.12.064

Effective protocols for the introduction of chiral α-trifluoromethyl benzyl moieties by nucleophilic displacement of enantiomerically enriched α-trifluoromethyl benzylic triflates are presented. The effects of substrate electronics, solvent polarity, temp

Full text of DOI:10.1016/j.tet.2008.12.064

Tetrahedron 65 (2009) 3189–3196
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Nucleophilic displacements of non-racemic a-trifluoromethyl benzylic triflates
*
Greg Hughes , Paul O’Shea, Julie Goll, Danny Gauvreau, Jennifer Steele
Department of Process Research, Merck Frosst Centre for Therapeutic Research, 16711 Route Transcanadienne, Kirkland, Que´bec, Canada H9H 3L1
a r t i c l e i n f o
a b s t r a c t
Article history:
Effective protocols for the introduction of chiral a-trifluoromethyl benzyl moieties by nucleophilic dis-
Received 10 September 2008
Received in revised form 16 December 2008
Accepted 18 December 2008
Available online 25 December 2008
placement of enantiomerically enriched
substrate electronics, solvent polarity, temperature, and base are studied by measuring the di-
astereomeric or enantiomeric excesses of the displacement products formed by coupling a variety of
trifluoromethyl benzylic triflates with a range of nucleophiles including amines, carboxylates, thiols, and
malonates. Preliminary investigations to elucidate the mechanism(s) involved in the loss of stereo-
chemical integrity at the benzylic center in the nucleophilic displacement reactions are also reported.
Ó 2008 Published by Elsevier Ltd.
a-trifluoromethyl benzylic triflates are presented. The effects of
a
-
Keywords:
Nucleophilic displacement
Trifluoromethyl
Benzylic cations
1. Introduction
displacements to highly electron deficient aromatic2 or hetero-
aromatic3 substrates where SN1 pathways would require highly
The nucleophilic displacement of a leaving group from benzylic
positions is a well studied classical transformation in organic
chemistry.1 While stereospecific SN2 displacement could be a useful
strategy for preparing optically active substrates, its use has been
limited due to the propensity for racemization via competing
SN1 pathways (Scheme 1). This tendency has limited such
destabilized carbocationic intermediates. The stereospecific SN2
displacement of particularly reactive diaryl carbinols following
activation at low temperature has also been reported.4
We were particularly interested in an alternative class of nu-
cleophilic displacements whereby carbocationic intermediates
would be destabilized by a benzylic CF3 substituent (compounds
SN1 Displacement
SN2 Displacement
LG
Nu
Nu
Nu
-LG
Nu
-LG
Ar
R
Ar
R
Ar
R
Ar = 4-CF3, R = CH3, LG = OTs
Ar = pyridyl, R = CH3, LG = OTs
Ar = Aryl, R = heteroaryl, LG = OTs
R = CH3
Ar = electron neutral or rich
Nucleophilic Displacement of Activated Aryl-CF3 Carbinols:
OSO2R
Nu
Nu
CF3
(+/-)B
Nu
or
Ar
CF3
Ar
CF3
Ar
A
B
Scheme 1. SN2 versus SN1 displacement of activated secondary benzylic alcohols.
A/B, Scheme 1). While the CF3 functionality in A would be
expected to retard SN1 processes, it is also well known to impede
SN2 inversions and there are very few examples of this type
of transformation being reduced to practice. Intramolecular
* Corresponding author. Tel.: þ1 732 594 5204; fax: þ1 732 594 0338.
E-mail address: greg_hughes@merck.com (G. Hughes).
0040-4020/$ – see front matter Ó 2008 Published by Elsevier Ltd.
doi:10.1016/j.tet.2008.12.064

3190
G. Hughes et al. / Tetrahedron 65 (2009) 3189–3196
displacements allowing access to CF3 substituted aziridines and
cyclopropanes have been reported.5 The displacement of tosylates
derived from fluoral hemiacetals with aluminate complexes has
also been shown to proceed with minimal (3–23%) racemization.6
During the course of detailed studies of the solvolysis of 1-aryl-
2,2,2-trifluoroethyl sulfonates, Tidwell et al. have shown that some
inversion occurred in the solvolysis of 1-phenyl-2,2,2-trifluroethyl
triflate in acetic acid.7 The strongest precedent for the trans-
formations we were interested in comes from the laboratories of
Fuchikami who demonstrated SN2 displacements of benzylic CF3-
mesylates using CsF as a base in DMF at 120 ꢀC with reaction times
of 3–4 days using benzoic acid (61% yield) as a nucleophile.8 The
only nitrogen nucleophile employed was phthalimide, which gave
only 11% yield. Milder conditions (60 ꢀC, 6 h) could be employed
using 5 equiv of thiophenol (71% yield) as a nucleophile. The same
this class of substrates with values ranging from ꢁ6.7 in TFA to
ꢁ11.9 in EtOH.7
2. Results
Our studies were initiated using mesylate 2a, derived from al-
cohol 1a with an ee of 90%.13 Employing the ethyl ester of (S)-leucine
(3a) as a nucleophile, we were unable to observe any of the desired
product using Fuchikami’s DMF/CsF protocol, with w40% of alcohol
1a being generated, presumably by attack of the amine at sulfur.
Therefore, we were pleased to find that the desired product 4 could
be obtained in 75% yield, albeit with partial racemization (10%) of
the benzylic stereocenter, by mixing 3a and triflate 2b in THF with
iPr2NEt as base (Table 1, entry 2).14 Switching to a more polar solvent
(acetonitrile) gave rise to a number of byproducts and more pro-
nounced racemization (24%, entry 3). While the conversions were
lower, similar yields and slightly higher diastereomeric excesses
were observed in iPrOAc, toluene, and n-Bu2O (70–78% yield, 6–7%
racemization, entries 4–6). The lower conversions in these solvents
could be compensated for by increasing the reaction temperature to
75 ꢀC. While more significant erosion in optical activity was ob-
served at this temperature, the desired product was formed in 89%
yield with a de of 72% (entry 7). Running the reaction in the absence
of solvent at 45 ꢀC further improved the de to 84% (entry 8). Alter-
natively, switching the base from iPr2NEt to K2CO3 in n-Bu2O at 75 ꢀC
improved theyieldandstereoselectivitysuchthat 4 couldbeisolated
in 94–96% yield and 83% de (entry 9). In order to simplify thework up
of the reactions, the solvent was switched to c-hexane, which
afforded the desired product in 98% yield and 86% de (entry 10).
With these conditions in hand, we turned our attention to
studying the influence of aromatic substituents on the stereo-
selectivity of the displacement reaction. A series of benzylic alco-
hols were subjected to the optimized SN2 displacement conditions
using 3b as a nucleophile (Table 2). A dramatic dependence on the
nature of the aromatic substituents was observed. While little or no
erosion was seen with electron deficient or electron neutral sub-
strates (0–11% erosion, entries 1–3), even a mildly electro-positive
substituent (CH3) gave significant erosion of optical activity at the
displacement center (59% racemization, entry 4). The use of a biaryl
benzylic alcohols resulted in extensive racemization, even with
a very electron deficient substrate (entry 5).
authors have also shown that optically active
a-trifluoromethyl
alkyl triflates could be displaced under milder conditions (rt to
60 ꢀC, 16–41 h) and in one case demonstrated the transformation to
be stereoselective, giving complete inversion of the stereocenter
using benzoic acid as a nucleophile. Unfortunately, no information
is available regarding the stereoselectivity of the benzylic mesylate
displacements.
As part of an ongoing preclinical program aimed at the de-
velopment of selective Cathepsin K inhibitors for the treatment of
osteoporosis,9 we have been interested in the synthesis of fluori-
nated
a-amino acids of general structure C (Scheme 2). While we
have previously reported an approach to the related
b
-amino al-
cohols D using the diastereoselective addition aryllithiates to op-
tically pure fluoral derived imines E,10 accessing the desired amino
acids would require a number of protecting group manipulations
and oxidation state changes. We have also developed a diastereo-
selective reductive amination between 2,2,2-trifluoroaceto-
phenones and
We felt that the direct SN2 displacement of a benzylic sulfonate
G with an -amino ester H would represent an attractive alternative
a
-aminoesters.11
a
as there are many proven methods for reducing aryl tri-
fluoromethyl ketones to provide optically active aryl-trifluoro-
methyl carbinols (I).12
We report herein the development of mild conditions, which
allow for the smooth SN2 displacement of either electronically
neutral or deficient secondary benzyl triflates with minimal erosion
of optical activity at the benzylic stereocenter. We have also ob-
served that even mildly electron donating substituents on the aro-
matic ring gives rise to a dramatic loss in optical activity at the
benzylic center. This is consistent with the results of Tidwell’s
The next issue to be addressed was the scope of the nucleo-
philes, which could be employed in the displacement reactions.
The reactions were run using 2b of >99.5% ee.15 Primary amines
such as amino diphenylmethane and a-methylbenzyl amine as well
as secondary amines such as morpholine underwent clean
solvolysis studies, which revealed exceptionally large
r
þ values for
Nucleophilic Displacement:
Diastereoselective Imine Addition:
CF3
CF3 R'
R'
CF3 R'
OH
OR''
OSO2R
N
H
H2N
OTES
R
N
R
O
O
R
H
H
D
G
C
CF3
OH
Li+
F3C
R'
R
OTES
R'
R
I
N
E
CF3
OH
OMe
O
R
OTES
H2N
F3C
F
Scheme 2. Accessing fluorinated a-amino acids via nucleophilic displacement.

G. Hughes et al. / Tetrahedron 65 (2009) 3189–3196
3191
Table 1
Optimization of conditions for nucleophilic displacement of benzylic sulfates
Me
Me
CO2Et
Me
Me
CO2Et
CF3
OSO2R
CF3
CF3
H2N
MeSO2Cl, pyridine
or, Tf2O, 2,6-Lutidine
Br
3a
OH
N
H
Base, Solvent
Temperature
20h
Br
Br
R = CH3, 2a
R = CF3, 2b
90% ee
(S),(S)-4
1a
Entry
Substratea
Solvent
Base
Temp (ꢀC)
Time (h)
% Conv
%dec
(% yield)b
1
2
3
4
5
6
7
8
9
10
2a
2b
2b
2b
2b
2b
2b
2b
2b
2b
DMF
THF
CsF
120
45
45
45
45
45
75
45
75
75
18
36
36
36
36
36
18
18
18
18
40 (0)d
96 (75)
85 (47)
83 (78)
83 (74)
78 (70)
100 (89)
88 (66)
100 (95)
100 (98)
d
iPrNEt2
iPrNEt2
iPrNEt2
iPrNEt2
iPrNEt2
iPrNEt2
iPrNEt2
K2CO3
K2CO3
80
66
83
84
84
72
85
83
86
CH3CN
iPrOAC
Toluene
n-Bu2O
n-Bu2O
None
n-Bu2O
c-Hexane
a
All of the starting materials were 90% ee.
Determined by calibrated HPLC analysis.
b
c
Measured by 19F NMR spectroscopic analysis.
d
Compound 1 (w40%) was generated during the reaction.
displacement with minimal erosion of % ee under the optimized
displacement conditions (Table 3, entries 1–4). We were surprised
to find that using the K2CO3/c-hexane protocol with thiophenol as
a nucleophile afforded only a moderate yield of the displacement
product with significant erosion of ee (45% yield, 62% ee, entry 5).
However, using the solvent free iPr2NEt protocol at 0 ꢀC, the desired
product could be isolated in 95% yield and 94% ee (entry 6). Similar
behavior was observed with octanoic acid as a nucleophile. Using
K2CO3 in c-hexane, the desired product was isolated in 75% yield
and 82% ee (entry 7). However, the solvent free iPr2NEt protocol at
45 ꢀC afforded the desired compound in 86% yield and 95% ee
(entry 8). It was also possible to form carbon–carbon bonds by
displacement with malonate anions at ambient temperature
without any erosion of ee (entry 9).
followed by stereospecific SN2 displacement; Ib The racemization
of triflate via an SN2 process, followed by stereospecific SN2 dis-
placement; II Competing SN1 and SN2 displacement with the nu-
cleophilic coupling partner; IIIa Base mediated epimerization of
triflate; IIIb Deprotonation of the triflate, followed by a-elimination
and N–H insertion of the resulting carbene; IIIc Base mediated
epimerization of the product.
The base mediated pathways IIIa–c are unlikely as they
would be expected to lead to more pronounced racemization of
electron deficient substrates. Additionally, upon reaching 100%
conversion, no further racemization of the displacement prod-
ucts was observed. Furthermore, stirring the triflates in the
presence of base at elevated temperatures in the absence of
nucleophiles actually retards the rate of racemization (and
chemical degradation) compared to a solution, which is warmed
in the absence of base.
3. Discussion
With displacement reactions involving electron deficient sub-
strates such as 2b, the % de/ee of the product decreases over the
course of the reaction, but at a rate, which is slower than the ero-
sion of % ee of residual triflate. For example, coupling 3a with 2b
There are a number of possible mechanisms by which the ra-
cemization observed in this study could be occurring (Scheme 3).
They include: Ia The racemization of triflate via an SN1 process
Table 2
Substituent effects on the displacement of 1-phenyl-2,2,2-trifluoroethyl triflates
CF3 CH3
CF3 CH3
CF3
1) Tf2O, 2,6-Lutidine
2)
N
3b (1.5 equiv)
OH
N
H
H
K2CO3 (1.6 equiv)
c-hexanes (1M), 75 °C
R
R
R
1a-e
(S),(S)-5a-e
(R),(S)-5a-e
Entry
Starting material
R
% eea
Yieldb
% dec
% Racemization
1
2
3
1a
1b
1c
1d
1e
CF3
Br
H
CH3
83
>99
92
86
>99
66
95
79
68
65
83
94
82
35
56
0
6
11
59
44
4
5d,e
4-MeSO2Ph
a
Measured by HPLC analysis of 1a–e.
b
c
HPLC.
Measured by 19F NMR spectroscopic analysis.
d
e
Compound 3b (4 equiv) was employed in a one-pot process.
DCE was used as solvent, and the reaction was run at 23 ꢀC.

3192
G. Hughes et al. / Tetrahedron 65 (2009) 3189–3196
Table 3
Scope of the nucleophile in the SN2 displacement of
a
-trifluoromethyl benzyl triflates
4, Nu = (S)-HN(iBu)CHCO2Et
5a, Nu = (S)-HN(Me)CHPh
6, Nu = N(CH2CH2)2O
7, Nu = HNCHPh2
CF3
OTf
CF3
NuH (3a-g)
Nu
Solvent, Temp
Base
Br
8, Nu = SPh
2b
Br
9, Nu = OCOHept
10, Nu = CH(CO2Me)2
1.0 equiv
3a, Nu = (S)-HN(iBu)CHCO2Et
3b, Nu = (S)-HN(Me)CHPh
3c, Nu = N(CH2CH2)2O
3d, Nu = HNCHPh2
3e, Nu = SPh
3f, Nu = OCOHept
3g, Nu = CH(CO2Me)2
4, 5a-10
>99.5 % ee
Entry
NuH
Product
Solvent
Base
Temp (ꢀC)
Time (h)
Yield (%)a
deb/eec (%)
1
2
3
4
5
6
7
8
9
3a
3b
3c
3d
3e
3e
3f
4
5a
6
7
8
8
9
9
10
c-Hexane
c-Hexane
c-Hexane
c-Hexane
c-Hexane
None
c-Hexane
None
THF
K2CO3
K2CO3
K2CO3
K2CO3
K2CO3
iPr2NEt
K2CO3
iPr2NEt
NaH
75
75
75
75
50
0
75
45
23
12
6
6
6
5
0.5
18
2
94
90
83
82
45
95
75
86
82
95
97
96
93
62
98
82
95
>99.5
3f
3g
18
a
Isolated yields.
b
c
Determined by 19F NMR spectroscopic analysis.
Measured by chiral HPLC analysis.
(90% ee) in neat iPr2NEt at 35 ꢀC gave 65% conversion after 6 h (Eq.
1). The product formed at this point had a de of 89%, while the
residual triflate had an ee of 87%. After 18 h and 87% conversion, the
product had a de of 87% and the residual triflate had an ee of 76%.
These results suggest that the SN1 pathway II is not playing
a dominant role in the racemization of this type of substrate, as one
would expect the % de of the product to be lower than the % ee of
the residual triflate if that were the case.
led us to speculate that solvated triflate ion might be undergoing
SN2 displacement of the starting triflate.
As further evidence of the effects of solvated triflate, a series of
reactions between 2b and 3b were run in DCE at 75 ꢀC using either
DIPEA or TMP as base with various amounts of added DIPEA$HOTf.
After >99% conversion, the % de of the product was measure by 19F
NMR and the results are represented graphically in Figure 1. With
no added DIPEA$HOTf, there was 17% erosion of de with TMP and
a loss of 32% with DIPEA. As the amount of added DIPEA$HOTf was
increased to a full equivalent, the extent of erosion increased to 45
and 50%, respectively.
The mechanism of racemization of more reactive substrates
appears to be different. In contrast to the displacement of 2b de-
scribed above, the biphenylsulphone derived triflate of 11 un-
dergoes displacement to give a product 12 whose % de lowers very
little throughout the course of the reaction. In addition, the residual
triflate at partial conversion was found to have a higher % ee than
the % de of the displacement product. For example, after 72 h at
25 ꢀC in t-BuOMe (TBME), HPLC showed 80% conversion. The
product formed to this point had a de of 67%, while the % ee of the
residual triflate had only dropped from >99.5% to 78% as de-
termined by chiral HPLC analysis (Eq. 2). These observations sug-
gest that with substrates, which are better able to stabilize benzylic
cations, there is a significant portion of the erosion of stereo-
chemical integrity, which can be attributed to SN1 displacement.
CF3
CF3
OTf
3a (1.5 equiv)
iPr2NEt (1.6 equiv), 35 °C
N
H
CO2Et
(1)
Br
Br
Time (h) Conv (%)
% ee 2b % de 4a
0
6
0
35
87
90
89
87
-
87
76
18
While the presence of SN1 type racemization of triflate (path Ia)
cannot be ruled out completely, the following observations lead us
to believe that electron deficient substrates racemize pre-
dominantly via path IIb. During an initial screen of solvents and
bases for the displacement reactions, a correlation was noted be-
tween the solubility of the triflate salts of the bases employed and
extent of erosion of optical activity (Table 4). For instance, using
dichloroethane (DCE) as solvent, the product formed upon coupling
2b with 3b was produced in higher % de with K2CO3 (89%, Table 4,
entry 1) and 2,2,6,6-tetramethylpiperidine (TMP) (83%, Table 4,
4. Conclusion
We have described rare examples of practical intermolecular
SN2 displacements of a leaving group positioned
a to a CF3 sub-
stituent.16 We have shown that a wide range of nucleophiles in-
cluding nitrogen, oxygen, and carbon based examples undergo
displacement in good yields and with excellent stereoselectivity
with electron deficient substrates. A strong correlation between the
electronic properties of the electrophilic partners and the stereo-
selectivity of the displacements has been established, and mecha-
nistic rationale for different modes of racemization for electron
poor and electron rich substrates has been discussed.
i
entry 2) as base than with either Pr2NEt (68%, Table 4, entry 3) or
2,6-lutidine (64%, Table 4, entry 4). The conjugate acids of the first
two bases (KOTf and TMP$HOTf) were highly insoluble, where as
i
the conjugate acids of 2,6-lutidine and Pr2NEt (2,6-lutidine$HOTf
and iPr2NEt$HOTf) have considerable solubility in DCE at 70 ꢀC. This

G. Hughes et al. / Tetrahedron 65 (2009) 3189–3196
Ib SN2 Racemization of triflate
3193
Ia SN1 Racemization of triflate
CF3
OTf
CF3
TfO
CF3
OTf
CF3
Nu
CF3
OTf
TfO
CF3
OTf
Nu
Ar
Ar
Ar
Ar
Ar
Ar
CF3
Nu
Nu
Ar
II SN1 Displacement
CF3
CF3
CF3
TfO
Nu
Ar
OTf
Ar
Nu
Ar
IIIb Deprotonation of triflate followed by α-elimination
IIIa Base mediated epimerization of triflate
CF3
CF3
OTf
CF3
Nu
CF3 Nu-H
C
CF3
OTf
CF3
Nu
H
B
F3C
Ar
Nu
OTf
B
H
F3C
Ar
H
OTf
Ar
Ar
Ar
Ar
OTf
Ar
insertion
Ar
IIIc Base mediated epimerization of displacement products
CF3
OTf
H
CF3
Nu
Scheme 3. Possible mechanisms of racemization.
CF3
F3C
Ar
B
Nu
H
Nu
Ar
Ar
Ar
Nu
purchased from Aldrich in Sure/SealÔ bottles and used as received.
Diisopropylethylamine (DIPEA), 2,2,6,6-tetramethylpiperidine
(TMP), and K2CO3 were purchased from Aldrich and used as re-
ceived. Unless otherwise noted, 1H NMR (500 MHz), 13C NMR
(125 MHz), and 19F NMR spectra (376.5 MHz) were recorded in ac-
etone-d6, chloroform-d1, or benzene-d6 purchased from CDN iso-
Table 4
Influence of base on the stereoselectivity of the displacement reactions
CF3 CH3
CF3
OTf
3b (1.5 equiv)
N
H
Ph
Base (1.6 equiv)
DCE, 75 °C
Br
4a
topes. The 1H and 13C spectra were referenced to residual solvent.19
F
2b
Br
>99.5% ee
NMR spectra were referenced to added benzotrifluoride
(ꢁ67.73 ppm). Coupling constants are reported in hertz (Hz). Mul-
tiplicities are as follows: s¼singlet, d¼doublet, t¼triplet, q¼quartet.
Chiral SFC was run on a Berger SFC system, reverse phase HPLC’s
were run on an Agilent 1100 series system. Infrared (IR) spectrawere
recorded on an Applied Systems Inc. ReactIR 1000, optics model.
Optical rotations were obtained on a Perkin Elmer 241 polarimeter.
Melting points are uncorrected. High resolution mass spectrometry
was performed by Merck Analytical Research. Elemental analysis
was performed by Prevalere Life Sciences in Whitesboro, NY.
Entry
Base
Conversiona (%)
Yielda (%)
deb (%)
1
2
3
4
K2CO3
TMP
DIPEA
2,6-Lutidine
>99
>99
>99
>99
94
88
59
82
89
83
68
64
a
Measured by calibrated HPLC analysis.
b
Measured by 19F NMR spectroscopic analysis.
5. Experimental
5.1. General
5.2. General procedure for triflation of aryl-2,2,2-
trifluoroethanols
Unless otherwise noted all reactions were run under an inert
atmosphere, and solvents and reagents were transferred by syringe.
Anhydrous DMF, cyclohexane, 1,2-dichloroethane (DCE), and iso-
propylacetate (IPAC) were purchased from A&C Chemicals Ltd, an-
hydrous grade methanol was purchased from Acros, and anhydrous
acetonitrile, tetrahydrofuran (THF), 1,2-dimethoxyethane (DME),
n-butyl ether (DBE), and tert-butyl methyl ether (MTBE) were
5.2.1. (1R)-1-(4-Bromophenyl)-2,2,2-trifluoroethyl
trifluoromethanesulfonate (2b)
Trifluoromethanesulfonic anhydride (1.43 mL, 8.52 mmol) was
added to
a
ꢁ15 ꢀC solution of 1R-(4-bromophenyl)-2,2,2-tri-
fluoroethanol 1a (1.45 g, 5.68 mmol, >99.5% ee, Chiralcel OJ
4.6 mmꢂ25 cm column; eluants: A, 2-propanol; B, CO2; 2 mL/min;
Racemization as a function of triflate concentration
55
50
45
40
35
30
25
20
15
DIPEA
TMP
0
0.1
0.2
0.3
0.4
Equivalents of added DIPEA.HOTf
0.5
0.6
0.7
0.8
0.9
1
Figure 1. Influence of added DIPEA$HOTf on the extent of racemization.

3194
G. Hughes et al. / Tetrahedron 65 (2009) 3189–3196
CF3
OTf
CF3
3a (1.5 equiv)
iPr2NEt (1.6 equiv), 35 °C
N
CO2Et
H
11
MeO2S
12
MeO2S
(2)
Time (h) Conv (%) % Ee 11 % De 12
0
72
0
80
>99.5
78
-
67
gradient: A/B 1:99 for 4 min to 20:80 over 12.7 min;
l
¼220 nm;
1.60 mmol) in c-hexane (1 mL). After 30 min, the mixture was di-
luted with water (10 mL) and c-hexane (10 mL). The organic layer
was washed with brine, dried over MgSO4, filtered, and concen-
trated. The residue was dissolved in c-hexane (1 mL) and K2CO3
temperature 35 ꢀC; tR: (R)-2a¼10.2 min, (S)-2a¼10.9 min) and 2,6-
lutidine (1.05 mL, 9.08 mmol) in c-hexane (5.7 mL). After 30 min,
water (50 mL) and c-hexane (50 mL) were added. The organic layer
was washed with water (2ꢂ25 mL) and concentrated. The residue
was diluted with c-hexane to give a solution whose total mass was
5.2 g. Quantitative HPLC analysis showed the material to be 40 wt %
(95% yield). This solution was stable and can be frozen and stored
for up to 6 months without decomposition or erosion of % ee. Chiral
HPLC analysis of the crude material showed it to be >99.5% ee
(ChiralpakÒ AD, 100% hexanes, 1 mL/min, 25 ꢀC). Samples could be
chromatographed (5% EtOAc/hexanes) to obtain pure samples by
1H NMR, but the pure material decomposes within a few days at
ambient temperature and after a few weeks at ꢁ20 ꢀC. The triflate
also underwent racemization on silica gel and material, which was
>99.5% ee prior to chromatography was 80% ee after. 1H NMR
(221 mg, 1.6 mmol) and 3b (193 mL, 1.50 mmol) were added. The
reaction mixture was warmed to 75 ꢀC for 18 h, then cooled to
ambient temperature and partitioned between c-hexane and water.
The organic layer was dried over MgSO4, filtered, and concentrated.
The residue was purified by flash chromatography. 1H NMR
(400 MHz, acetone):
d
7.69 (d, J¼8 Hz, 2H), 7.63 (d, J¼8 Hz, 2H),
7.17–7.27 (m, 5H), 4.39 (quint, J¼8 Hz, 1H), 3.98 (quint, J¼6.5 Hz,
1H), 2.87 (t, J¼6.5 Hz, 1H), 1.37 (d, J¼6.5 Hz, 3H); 13C NMR
(126 MHz, acetone):
d
145.7, 141.0, 130.8 (q, J¼32 Hz), 130.1, 129.1,
127.8, 127.4, 127.1 (q, J¼237 Hz), 126.0 (q, J¼3.5 Hz), 124.9 (q,
J¼226 Hz), 62.0 (q, J¼28.5 Hz), 57.2, 23.7; 19F NMR (377 MHz, ace-
tone):
d
ꢁ67.73 (s, 3F), ꢁ78.14 (d, J¼7 Hz, 3F).
(400 MHz, acetone):
d
7.83–7.79 (m, 2H), 7.70 (d, J¼8.4 Hz, 2H), 6.74
(q, J¼5.9 Hz, 1H); 13C NMR (126 MHz, benzene):
d
132.6, 129.4,
5.4.2. (1S)-2,2,2-Trifluoro-1-phenyl-N-[(1S)-1-
phenylethyl]ethanamine
Alcohol 1c (92% ee) was subjected to the general conditions for
triflation/displacement to afford the title compound in 79% yield and
82%de. Thespectraldatawereconsistentwiththeliteraturevalues.17
126.9, 126.2, 121.7 (q, J¼282 Hz), 118.7 (q, J¼320 Hz), 82.0 (q,
J¼36 Hz); 19F NMR (377 MHz, acetone):
d
ꢁ80.33 (s, 3F), ꢁ81.77 (d,
J¼6 Hz, 3F); IR (neat, cmꢁ1) 1602, 1327, 1247, 1197, 1135, 1073.
5.3. General K2CO3/c-hexane displacement conditions
5.4.3. (1S)-2,2,2-Trifluoro-1-(4-methylphenyl-N-[(1S)-1-
phenylethyl]ethanamine) (5d)
5.3.1. (1S)-1-(4-Bromophenyl)-2,2,2-trifluoro-N-[(1S)-1-
phenylethyl]ethanamine (5a)
Alcohol 1d (86% ee, ChiralpakÒ AD, 3% isopropanol/hexanes,
1 mL/min, 25 ꢀC) was subjected to the general method for triflation/
displacement through process affording the title compound as
a colorless oil in 68% yield (35% de). 1H NMR (400 MHz, acetone):
Potassium carbonate (1.26 g, 9.08 mmol) and 3b (1.08 mL,
8.52 mmol) were added to 5.5 g of a 40 wt % solution of 2b
(5.68 mmol) in c-hexane and the mixture was warmed to 75 ꢀC and
aged for 18 h. The mixture was cooled to ambient temperature and
partitioned between cyclohexane (50 mL) and 1 M HCl (50 mL). The
organic layer was washed with water (2ꢂ25 mL). The organic layer
was concentrated to dryness. The residue was purified by flash
chromatography (5% ethylacetate/hexane) to afford the desired com-
pound as a colorless oil (1.83 g, 90%). 1H NMR (400 MHz, acetone):
d
7.20–7.30 (m, 7H), 7.17 (d, J¼8 Hz, 2H), 4.16 (quint, J¼8 Hz,1H, (S,S)
diastereomers); 13C (126 MHz, acetone):
146.0, 139.0, 133.3, 129.9,
129.1, 129.0, 127.8, 127.4, 127.8 (q, J¼282 Hz), 62.0 (q, J¼28 Hz), 56.7,
23.6, 21.1; 19F NMR (377 MHz, acetone):
ꢁ78.33 (d, J¼7 Hz).
d
d
5.4.4. (1S)-2,2,2-Trifluoro-1-(4-(4-methylsulfonylphenyl)phenyl-
N-[(1S)-1-phenylethyl]ethanamine) (5e)
d
1.41 (d, J¼6.6 Hz, 3H), 2.66 (t, J¼6.7 Hz, 1H), 4.02 (quint, J¼6.3 Hz,
1H), 4.28 (quint, J¼7.6 Hz, 1H), 7.19–7.24 (m, 1H), 7.25–7.32 (m, 4H),
Alcohol 1e (>99.5% ee) was subjected to the general method for
triflation/displacement through process, except that 1,2-di-
chloroethane (DCE) was used in place of c-hexane in order to improve
solubility. The displacement was also run at ambient temperature,
affording the title compound as a colorless oil in 68% yield (56% de). 1H
7.36 (d, J¼8.3 Hz, 2H), 7.52 (d, J¼8.4 Hz, 2H); 13C NMR (126 MHz,
acetone):
d
23.7, 56.9, 61.7 (q, J¼19.4 Hz), 122.9, 126.8 (q, J¼283 Hz),
127.3, 127.7, 129.0, 131.1, 132.2, 135.6, 145.6; 19F NMR (377 MHz, ace-
tone):
d
ꢁ78.37 (d, J¼7.5 Hz); IR (neat, cmꢁ1): 3350, 3030, 2968, 1594,
1490, 1366, 1258, 1119, 814, 702. Anal. Calcd for C16H15BrF3N: C, 53.65;
H, 4.22; N, 3.91. Found: C, 53.63; H, 4.28; N, 4.21.
NMR (500 MHz, acetone):
d
8.03(d, J¼8.5 Hz, 2H), 7.93 (d, J¼8.5 Hz, 2H),
7.74 (d, J¼8.3 Hz, 2H), 7.57 (d, J¼7.9 Hz, 2H), 7.35–7.19 (m, 5H), 4.35–4.29
(m, 1H), 4.02–3.96 (m, 1H), 3.16 (s, 3H), 1.36 (d, J¼6.6 Hz, 3H); 19F NMR
5.4. General method for triflation/displacement
through process
(377 MHz, acetone):
C, 63.73; H, 5.12; N, 3.23. Found: C, 63.77; H, 5.00; N, 3.12.
d
ꢁ78.05(d, J¼7 Hz). Anal. Calcdfor C23H22F3NO2S:
5.4.5. Ethyl 2-{[(1S)-1-(4-bromophenyl)-2,2,2-
5.4.1. (1S)-2,2,2-Trifluoro-N-[(1S)-1-phenylethyl]-1-[4-
(trifluoromethyl)phenyl]ethanamine (5b)
trifluoroethyl]amino}-4-methylpentanoate (4)
The general K2CO3/c-hexane procedure was employed. The
product was purified by flash chromatography (10% EtOAc/
Tf2O (252
m
L, 1.50 mmol) was added drop-wise to a ꢁ10 ꢀC so-
lution of 1b (244 mg, 1.00 mmol, 83% ee) and 2,6-lutidine (185 mL,

G. Hughes et al. / Tetrahedron 65 (2009) 3189–3196
3195
hexanes) to afford the desired compound as a colorless oil in 94%
yield. 1H NMR (400 MHz, acetone):
5.5.2. (1S)-1-(4-Bromophenyl)-2,2,2-trifluoroethyl phenyl
sulfide (8)
d
0.90 (d, J¼6.5 Hz, 3H), 0.92 (d,
i
J¼6.5 Hz, 3H), 1.10 (t, J¼7 Hz, 3H), 1.41–1.54 (m, 2H), 1.88 (nonet,
J¼6.5 Hz, 1H), 2.48 (dd, J¼9, 5.5 Hz, 1H), 3.43–3.50 (m, 1H), 3.90–
3.97 (m, 2H), 4.33–4.39 (m, 1H), 7.44 (d, J¼8.4 Hz, 2H), 7.58 (d,
The general protocol for neat Pr2NEt displacement reactions
was followed except that the reaction was carried out at 0 ꢀC for
30 min. The crude reaction mixture was purified to flash chroma-
tography (0/2% EtOAc/hexanes) to afford the desired compound
in 95% yield. Chiral HPLC analysis (SFC, ChiralpakÒ OD-H, 1% MeOH/
COs2c, hold 4 min, then to 20% MeOH/COs2c at 1.3%/min, hold 5 min,
2.0 mL/min, 220 nm, 50 ꢀC, retention times: 9.9 min, 10.5 min)
showed the product to be 98% ee. Mp (TBME)¼58–59 ꢀC; 1H NMR
J¼8.5 Hz, 2H); 13C NMR (126 MHz, acetone):
d 14.3, 22.1, 23.1, 25.3,
43.3, 60.2, 61.1, 63.5 (q, J¼29 Hz, 1H), 123.3, 126.6 (q, J¼281 Hz),
131.5, 132.4, 135.7, 175.3; 19F NMR (377 MHz, acetone):
d
ꢁ79.15 (d,
J¼8 Hz); IR (neat, cmꢁ1): 3343, 2961, 1729, 1490, 1262, 1151, 1123,
1011, 814. Anal. Calcd for C16H21BrF3NO2: C, 48.50; H, 5.34; N, 3.53.
Found: C, 48.51; H, 5.27; N, 3.62.
(CD3COCD3):
d
5.20 (q, J¼8.7 Hz, 1H), 7.33–7.35 (m, 3H), 7.44–7.50
(m, 4H), 7.59 (d, J¼8.5 Hz, 2H); 13C NMR (CDCl3):
55.1 (q, J¼29 Hz),
d
5.4.6. 4-[(1S)-1-(4-Bromophenyl)-2,2,2-trifluoroethyl]-
morpholine (6)
123.4, 126.6 (q, J¼279 Hz), 129.5, 130.1, 131.9, 132.7, 132.9, 133.9,
589
134.0; 19F NMR (CD3COCD3):
d
ꢁ72.96 (d, J¼11.5 Hz); [
a
]
þ261
D
The general K2CO3/c-hexane procedure was employed. The
product was purified by flash chromatography (0/10% EtOAc/
hexanes) to afford the desired compound as a colorless oil in 83%
yield. Chiral HPLC analysis (ChiralpakÒ AD, 1% IPA/CO2sc, hold 4 min,
then to 20% IPC/COs2c at 1.3%/min, hold 5 min, 1.5 mL/min, 220 nm,
50 ꢀC, retention times: 10.4 min, 11.0 min) showed the material to
(17.1 mg/mL, CHCl3). Anal. Calcd for C14H10BrF3S: C, 48.43; H, 2.90.
Found: C, 48.48; H, 3.05.
5.5.3. Dimethyl [(1S)-1-(4-bromophenyl)-2,2,2-
trifluoroethyl]malonate (10)
A solution of 2b (500 mg, 1.29 mmol) in c-hexane (1 mL) was
added to a rt mixture of NaH (103 mg of 60 wt % NaH, 2.58 mmol)
be 96% ee. 1H NMR (400 MHz, acetone):
d
7.63 (d, J¼8.5 Hz, 2H),
7.42 (d, J¼8.4 Hz, 2H), 4.35 (q, J¼9 Hz, 1H), 3.60 (t, J¼4.7 Hz, 4H),
and dimethyl malonate (221 mL, 1.94 mmol) in THF (5 mL). After
2.59–2.65 (m, 2H), 2.49–2.57 (m, 2H); 13C NMR (126 MHz, acetone):
warming to 23 ꢀC and aging 18 h, HPLC analysis showed 100%
conversion. The mixture was diluted with TBME (25 mL) and
quenched carefully with 1 N NaOH (10 mL). The organic layer was
washed with brine, dried over MgSO4, filtered, and concentrated.
The residue was purified by flash chromatography (5/20% EtOAc/
Hexanes) to afford the desired compound as a white solid (390 mg,
d
132.4, 132.3, 132.1, 126.7 (q, J¼284 Hz), 123.3, 70.2 (q, J¼27.5 Hz),
67.5, 51.9; 19F NMR (377 MHz, acetone):
d
ꢁ72.18 (d, J¼8.5 Hz); IR
(neat, cmꢁ1) 2964, 2856, 1490, 1455, 1254, 1150, 1188, 1011, 872,
726; [
589
a]
þ40 (c 11.3 mg/mL, CHCl3).
D
82%). Chiral HPLC analysis (Chiralpak OD-H, 1% PrOH/COs2c, hold
i
5.4.7. (1S)-1-(4-Bromophenyl)-N-(diphenylmethyl)-2,2,2-
trifluoroethanamine (7)
The general K2CO3/c-hexane procedure was employed. The re-
action mixture was purified by flash chromatography (0/5%
EtOAc/hexanes) to afford the title compound in 82% yield as a col-
4 min, then to 20% iPrOH at 1.3%/min, hold 5 min, retention times:
5.8 min, 6.6 min) showed >99.5% ee. Mp (TMBE): 101–102 ꢀC; 1H
NMR (CD3COCD3):
d
3.46 (s, 3H), 3.78 (s, 3H), 4.28 (d, J¼11.3 Hz,1H),
4.32–4.40 (m, 1H), 7.45 (d, J¼8.3 Hz, 2H), 7.59 (d, J¼8.5 Hz, 2H); 13C
orless oil. 1H NMR (400 MHz, acetone):
d
7.62 (d, J¼8.5 Hz, 2H), 7.45
NMR (C6D6):
d
49.2 (q, J¼27.5 Hz), 52.3 (br peak), 52.4, 52.8, 123.3,
(dd, J¼7, 5 Hz, 4H), 7.34 (dd, J¼16, 8 Hz, 4H), 7.52 (t, J¼7.5 Hz, 3H),
126.1 (q, J¼281 Hz),131.3,131.4,132.0,165.8,166.9; 19F (CD3COCD3):
7.18 (t, J¼7.5 Hz, 1H), 4.76 (d, J¼3.7 Hz, 1H), 4.17 (quint, J¼8 Hz, 1H),
d
81.26 (d, J¼6.9 Hz); IR (film, cmꢁ1) 1752, 1733, 1489, 1432, 1309,
589
3.22 (dd, J¼10, 3.5 Hz, 1H); 13C NMR (126 MHz, acetone):
d
144.4,
1161, 1112, 1077, 1011, 973, 930, 814; [
a
]
þ28 (c 13.6, CHCl3). Anal.
D
143.2, 134.8, 132.6, 131.6, 129.5, 129.2, 128.3, 128.2, 128.0, 127.9,
Calcd for C13H12BrF3O4: C, 42.30; H, 3.28. Found: C, 42.51; H, 3.36.
126.4 (q, J¼281 Hz), 123.4, 64.6, 62.1 (q, J¼29 Hz); 19F NMR
(377 MHz, acetone):
1258, 1170, 1123, 702; [
for C21H17BrF3N: C, 60.02; H, 4.08; N, 3.33. Found: C, 59.83; H, 4.34;
N, 3.37.
d
ꢁ78.49 (d, J¼8 Hz); IR (neat, cmꢁ1) 1490,
Acknowledgements
589
a]
þ89 (c 13.4 mg/mL, CHCl3). Anal. Calcd
D
The authors thank Dr. Chris Bayly and Dr. Daniel MacKay for
helpful suggestions, and Dr. Krista Morley and Dr. Mattew Tudge for
synthetic assistance. J.G. and J.S. thank NSERC for industrial un-
dergraduate research fellowships.
i
5.5. General neat PrNEt displacement reactions
5.5.1. (1S)-1-(4-Bromophenyl)-2,2,2-trifluoroethyl octanoate (9)
References and notes
Octanoic acid (307
of 2b (500 mg, 1.29 mmol) and Pr2NEt (368
m
L, 1.94 mmol) was added to a 25 ꢀC mixture
i
m
L, 2.06 mmol). The
1. (a)Allen, A.D.;Kanagasabapathy, V.M.;Tidwell,T. T.J.Am. Chem.Soc.1985,107, 4513;
(b) Richard, J. P.;Jencks,W. P. J. Am. Chem.Soc.1984,106,1383 and references therein.
2. Lim, C.; Kim, S.-H.; Yoh, S.-D.; Fujio, M.; Tsuno, Y. Tetrahedron Lett. 1997, 38, 3243.
3. Uenishi, J.; Hamada, M.; Aburatani, S.; Matsui, K.; Yonemitsu, O.; Tsukube, H.
J. Org. Chem. 2004, 69, 6781.
4. Bolshan, Y.; Chen, C.; Chilenski, J. R.; Gosselin, F.; Mathre, D. J.; O’Shea, P. D.; Roy,
A.; Tillyer, R. D. Org. Lett. 2004, 6, 111.
5. (a) Katagiri, T.; Ihara, H.; Takahashi, M.; Kashino, S.; Furuhashi, K.; Uneyama, K.
Tetrahedron: Asymmetry 1997, 8, 2933; (b) Katagiri, T.; Irie, M.; Uneyama, K.
Tetrahedron: Asymmetry 1999, 10, 2583.
6. (a) Matsutani, H.; Poras, H.; Kusumoto, T.; Hiyama, T. Synlett 1998, 1353; (b)
Matsutani, H.; Kusumoto, T.; Hiyama, T. Chem. Lett. 1999, 529.
7. Allen, A. D.; Ambidge, A. I.; Che, C.; Micheal, H.; Muir, R. J.; Tidwell, T. T. J. Am.
Chem. Soc. 1983, 105, 2343.
8. (a) Hagiwara, T.; Tanaka, K.; Fuchikami, T. Tetrahedron Lett. 1996, 37, 8187; (b)
Hagiwara, T.; Ishizuka, M.; Fuchikami, T. Nippon Kagaku Kaishi 1998, 11, 750.
9. (a) Gauthier, J. Y.; Chauret, N.; Cromlish, W.; Desmarais, S.; Duong, L. T.; Fal-
gueyret, J.-P.; Kimmel, D. B.; Lamontagne, S.; Leger, S.; LeRiche, T.; Li, C. S.; Masse,
F.; McKay, D. J.; Nicoll-Griffith, D. A.; Oballa, R. M.; Palmer, J. T.; Percival, M. D.;
Riendeau, D.; Robichaud, J.; Rodan, G. A.; Rodan, S. B.; Seto, C.; Therien, M.;
Truong, V.-L.; Venuti, M. C.; Wesolowski, G.; Young, R. N.; Zamboni, R.; Black, W. C.
Bioorg. Med. Chem. Lett. 2008, 18, 923; (b) Black, W. C.; Bayly, C. I.; Davis, D. E.;
Desmarais, S.; Falgueyret, J.-P.; Le´ger, S.; Li, C. S.; Masse´, F.; McKay, D. J.; Palmer, J.
mixture was warmed to 45 ꢀC for 2 h. HPLC analysis showed 100%
conversion. The mixture was diluted with TBME (25 mL) and
washed with 1 N HCl (25 mL) and brine. The organic layer was dried
over MgSO4, filtered, and concentrated to dryness. The residue was
purified by flash chromatography (100% hexanes) to afford the
desired product as a colorless oil (423 mg, 86%). Chiral HPLC anal-
i
ysis (SFC, ChiralpakÒ AD, 1% PrOH/COs2c, hold 4 min, then to 20%
iPrOH at 1.3%/min, hold 5 min, 220 nm, 2.0 mL/min, retention
times: 5.4 min, 8.7 min) showed the product to be 95% ee. 1H NMR
(CD3COCD3):
d
0.86 (t, J¼7.0 Hz, 3H), 1.22–1.34 (m, 8H), 1.59–1.68
(m, 2H), 2.53 (t, J¼7.5 Hz, 2H), 6.33 (q, J¼7 Hz,1H), 7.54 (d, J¼8.4 Hz,
2H), 7.66 (d, J¼8.4 Hz, 2H); 13C NMR (CDCl3):
d 14.0, 22.5, 24.7, 28.8,
28.9, 31.6, 33.9, 71.1 (q, J¼33 Hz), 122.9 (q, J¼280 Hz), 124.2, 129.6,
130.4, 131.9, 171.4; 19F NMR (377 MHz, CD3COCD3):
d
ꢁ73.10 (d,
J¼8.2 Hz); IR (neat, cmꢁ1) 2930, 1760, 1355, 1266, 1181, 1135, 1011,
589
811; [
a
]
D
þ63 (c 11.8 mg/mL, CHCl3). Anal. Calcd for C16H20BrF3O2:
C, 50.41; H, 5.29. Found: C, 50.39; H, 5.17.

3196
G. Hughes et al. / Tetrahedron 65 (2009) 3189–3196
T.; Percival, M. D.; Robichaud, J.; Tsou, N.; Zamboni, R. Bioorg. Med. Chem. Lett.
2005,15, 4741; (c) Li, C. S.; Deschenes, D.; Desmarais, S.; Falgueyret, J.-P.; Gauthier,
J. Y.; Kimmel, D. B.; Le´ger, S.; Masse´, F.; McKay, D. J.; Percival, M. D.; Riendeau, D.;
14. That the nucleophilic displacement occurs with inversion was determined by
hydrolyzing ester (S),(S)-4a to the corresponding carboxylic acid and compar-
ing its spectral data to the literature values. Roy, A.; O’Shea, P. D.; Cheng, C.-y.
J. Org. Chem. 2006, 71, 4320. Authentic samples of all four stereoisomer were
prepared and HPLC analysis confirms that the racemization in this process is
occurring at the benzylic stereocenter with the stereochemical integrity of the
epimerizable
a-amino ester stereocenter remaining intact.
15. Alcohol 1a formed in 90% ee was recrystallized from hexanes with 80% re-
covery as a single isomer as determined by chiral HPLC analysis.
´
Sevgi, B. R.; McGrath, M. E.; Therien, M.; Truong, V.-L.; Wesolowski, G.; Zamboni,
R.; Black, W. C. Bioorg. Med. Chem. Lett. 2006, 16, 1985.
10. Gosselin, F.; Roy, A.; O’Shea, P. D.; Chen, C.; Volante, R. P. Org. Lett. 2004, 6, 641.
11. Hughes, G.; Devine, P. N.; Naber, J. R.; O’Shea, P.; Foster, B.; McKay, D. J.; Volante,
R. P. Angew. Chem., Int. Ed. 2007, 46, 1839.
12. (a) Corey, E. J.; Bakshi, R. K. Tetrahedron Lett. 1990, 31, 611; (b) Fujisawa, T.;
Ichikawa, K.; Shimizu, M. Tetrahedron: Asymmetry 1993, 4, 1237; (c) Matsuda, T.;
Harada, T.; Nakajima, N.; Itoh, T.; Nakamura, K. J. Org. Chem. 2000, 65, 157.
13. The benzylic alcohols were prepared by the oxazaborolidine ((S)-Bu-OAB) cata-
lyzed catacholborane reduction of the corresponding ketones described in Ref.12a.
16. As a testament to the robustness of this methodology, the displacement of 2b
with an a-amino ester has been performed on 2 kg scale, affording the desired
product in 94% yield with only 3% racemization of the benzylic stereocenter.
17. Pirkle, W. H.; Hauske, J. R. J. Org. Chem. 1977, 42, 2436.