Stereodivergent synthesis of novel chiral C2-symmetric bis-trifluoromethyl-2-oxazolidinones

Article abstract of DOI:10.1016/j.tetlet.2018.04.072

A simplified effective synthetic process is described for the diastereoselective synthesis of the chiral C₂-symmetric CF₃-ureas (R,R)-15 and (S,S)-15 from (S)-α-phenylethylamine, glyoxal and CF₃I.

Full text of DOI:10.1016/j.tetlet.2018.04.072

Accepted Manuscript
Stereodivergent Synthesis of Novel Chiral C 2-Symmetric Bis-Trifluorometh-
yl-2-oxazolidinones
Travis Remarchuk, E.J. Corey
PII:
S0040-4039(18)30558-6
https://doi.org/10.1016/j.tetlet.2018.04.072
TETL 49935
DOI:
Reference:
Tetrahedron Letters
To appear in:
Received Date:
Revised Date:
Accepted Date:
28 February 2018
25 April 2018
26 April 2018
Please cite this article as: Remarchuk, T., Corey, E.J., Stereodivergent Synthesis of Novel Chiral C 2-Symmetric
Bis-Trifluoromethyl-2-oxazolidinones, Tetrahedron Letters (2018), doi: https://doi.org/10.1016/j.tetlet.
2018.04.072
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Graphical Abstract
Stereodivergent Synthesis of Novel Chiral
C2-Symmetric Bis-Trifluoromethyl-2-
oxazolidinones
Travis Remarchuk†a and E. J. Corey†*
Leave this area blank for abstract info.

1
Tetrahedron Letters
Stereodivergent Synthesis of Novel Chiral C2-Symmetric Bis-Trifluoromethyl-2-
oxazolidinones
Travis Remarchuk†a and E. J. Corey†
† Department of Chemistry and Chemical Biology, Harvard University, Cambridge, Massachusetts 02138, United States
ARTICLE INFO
ABSTRACT
Article history:
Received
Received in revised form
Accepted
A simplified effective synthetic process is described for the diastereoselective synthesis of the
chiral C2-symmetric CF3-ureas (R,R)-15 and (S,S)-15 from (S)--phenylethylamine, glyoxal and
CF3I.
2009 Elsevier Ltd. All rights reserved.
Available online
Keywords:
Nucleophilic Trifluoromethylation
Stereoselective
Chiral 2-oxazolidinone
Vicinal Diamine
[2+1]-Cycloaddition
to 1-heptyne (or other terminal acetylenes), using a variety of
 Stereoselective double trifluoromethylation
reaction of chiral bis-imines
 Stereodivergent synthesis of C2-symmetric
chiral 2-oxazolidinones
 Enantioselective [2+1]-cycloaddition with
mixed-ligand dirhodium complexes
chiral amidate-bridged dirhodium (II) complexes of the type
Rh2(RCO2)n(L*4–n).1 In the case of 1-heptyne, ethyl (1S)-2-n-
amyl-2-cyclopropenylcarboxylate 1 was formed in >90% ee with
the chiral catalysts 2–5 which contained the chiral ligands (R,R)-
di-tert-butyl-N-triflylimidazolidinone (DTBTI) 6, or (R,R)-
diphenyl-N-triflylimidazolidinone (DPTI) 7 (Figure 1). Catalysts
2 and 4, which have the M-cis-anti configuration,2 gave 91% ee
of the cyclopropene product, and the cis-2/1 catalyst 5 gave even
higher ee (up to 95% ee).
From these and other results, we surmised that both steric bulk
of the substituent alpha to N and lower electron density of the
Several years ago, we described
e—na—ntio—selective [2+1]-cycloaddition of ethyl diazoacetate (EDA)
a detailed study of
 Corresponding author. Tel.: +1-617-495-4033; fax: +1-617-495-0376; e-mail: corey@chemistry.harvard.edu
a Present address: Denali Therapeutics Inc., South San Francisco, CA 94080

2
Tetrahedron
CF3I (6.4 equiv)
Zn (10 equiv)
DMF
nitrogen ligand are favorable to enantioselectivity. The
Ph
Ph
Ph
Ph
Ph
Ph
Ph
calculated pKa value of the DPTI ligand 7 is 9.5, whereas that of
the DTBTI ligand 6 is 10.8.3 In an effort to design a 2-
imidazolidinone ligand with a lower pKa than DPTI, while
maintaining much of the bulk of the tert-butyl group in DTBTI,
we focused on the trifluoromethyl (CF3) analog. The calculated
pKa of the analogous di-trifluoromethyl-N-triflylimidazolidinone
(DTFTI) derivative is 6.6, and sterically the CF3 group is
reported to be similar in size to an isopropyl group, having a Van
der Waals’ volume (hemisphere) of 42.6 Å3 (relative to 16.8 Å3
for CH3).4
NH HN
F3C CF3
N
N
Ph
–35 °C, 36 h
8
(31%)
(±)-9
(X-ray)
ORTEP of (±)-9
Under similar reaction conditions described above, the reaction
with bis-imine 11 again resulted in a mixture of mono- and bis-
CF3-addition products after 24 h reaction time. Purification by
silica gel chromatography resulted in a 1:1.5 mixture of C2-
symmetric diastereomers in 32% yield.12 The two products, later
assigned as (S,S,S,S)-12 and (S,R,R,S)-13 respectively, had one
diagnostic 19F NMR signal each, thus also supporting the C2-
symmetry.
Optimization studies were performed with bis-imine 11 and
several key variables were identified. Highest yields were
obtained using DMF solvent, presumably due to the stabilized
CF3ZnI-DMF complex that is formed. The use of freshly
activated Zn13 was helpful since the reaction of Zn and CF3I is
highly exothermic, and rapid initiation is critical. An initial
temperature of ca. –50 °C for addition of CF3I and then
maintaining the reaction temperature between –45 and –25 °C
was optimal. Initiating the reaction at 0 °C or above have led to
complex reaction mixtures and no product formation. Substrate
concentration between 0.1–0.8M gave similar results with the dr
of 12/13 between 1:1.1–1.5. More concentrated mixtures often
resulted in complex, intractable mixtures of products, whereas
lower concentrations led to slow reaction and stalling. Additives
did not lead to improved dr or reaction yields, including the
following: I2 (>1 equiv), ZnCl2 (1 equiv), NaI (5 equiv), LiI (1
equiv), and TiCl4 (1 equiv).
Figure 1. Enantioselective cyclopropenation reactions with mixed-ligand
Rh2(II) catalysts 2–5.
Surprisingly, at the onset of this work, there were no reports of
the synthesis of the required chiral C2-symmetric bis-
trifluoromethylethylenediamine, the required precursor to the 2-
oxazolidinone of DTFTI, probably due to the challenge of
introducing trans vicinal CF3 groups stereoselectively.5 Herein,
we report studies directed at the synthesis of the required
The optimal reaction conditions found to date involve using up to
8 equiv of activated Zn and 8 equiv CF3I relative to bis-imine 11
in DMF cooled to –30 °C. The reaction mixture is maintained
hexafluorodiamines and the corresponding cyclic ureas.6
Generation of highly reactive nucleophilic sources of
trifluoromethyl such as CF3Li and CF3Mg are not practical to use
due to their instability,7 however, the CF3ZnI complex has been
reported and used for the trifluoromethylation of carbonyl
compounds.8 This complex is formed in-situ by reaction of Zn
with trifluoromethyl iodide (CF3I, bp –23 °C) in the presence of
the aprotic donor solvent DMF. Our first successful
trifluoromethylation reaction was with this reagent and the
simple achiral benzhydryl-substituted bis-imine 8.9 The reaction
of CF3I with a cooled mixture of Zn, 8, and DMF, produces an
orange to deep red mixture and after a period of 36 h, the starting
material was mostly consumed as indicated by TLC analysis. The
desired 1,2-trans CF3 diamine (±)-9 was isolated in 31% yield
after chromatography (Scheme 2), the stereochemistry being
confirmed by X-ray crystallographic analysis. Further
optimization studies with bis-imine 8 did not lead to improved
reaction yields.10 Because it was problematic to install both CF3
groups efficiently, and because resolution would be required, we
focused on using the chiral bis-imine 11 derived from (S)--
phenylethylamine (Scheme 4).11
between –25 and –20 °C for a period of 24 h under an inert
atmosphere. After workup and purification by chromatography, a
~1:1.2 mixture of (S,S,S,S)-12 and (S,R,R,S)-13 can be obtained
in yields up to 65%.14
The lack of diastereoselectivity in the addition to 11 suggests the
possibility that the first addition is non-enantioselective and that
the second reaction proceeds via a Zn(II) chelate (I in Scheme 4).
Also of mechanistic interest is the fact that only the (S,R,R,S) and
(S,S,S,S) diastereomers are formed and that none of the meso
isomer is detected. This result is contrary to high
diastereoselectivity observed in similar systems and in the
absence of the strong donor solvent DMF that are governed by
chelate control.15 Thus, the first CF3 addition to 11 is not 1,3-
stereoselective with respect to the chiral auxiliary, whereas the
addition of second CF3 group occurs with high 1,2-
stereoselectivity. By following the reaction using 19F NMR
spectroscopy and an internal standard PhCF3 (–63.5 ppm), we
have observed that the reaction of Zn and CF3I (–12 ppm) gives a
ca. 1:1 mixture of CF3ZnI (–44.6 ppm) and (CF3)2Zn (–43.1 ppm,
s), confirming previous reports.16 The CF3ZnI-DMF complex
(–77 ppm, d, J = 4.5 Hz) can also sometimes be detected at the
reactions temperatures, but can lead to formation of fluoral
Scheme 3. Double trifluoromethylation of bis-imine 8.

3
(CF3CHO) (–84 ppm) at warmer temperatures. Protic sources can
also lead to the formation of fluoroform (–80 ppm, d, J = 79 Hz).
crystalline 16, (mp 129 °C, onset by DSC). The structure of 16
(DTFTI) was confirmed by X-ray crystallographic analysis.
The formation of the less reactive (CF3)2Zn is unavoidable using
this procedure, and necessitates the use of 6-8 equiv of CF3I in
the reaction. A control experiment in which the (CH3)2Zn reagent
was added to a DMF solution of 11 cooled to –40 °C resulted in
diimine decomposition and no dimethyl addition product. In
another control experiment to test if radicals are involved, the
radical scavenger 2-methyl-2-butene (8 equiv) was added to the
reaction mixture with 8 equiv of both CF3I and Zn relative to bis-
imine 11. There was no impact to the trifluoromethylation
reaction, and a 1:1.2 mixture of 12/13 was still produced in ~60%
yield.
Scheme 6. Synthesis of DTFTI ligand (R,R)-16.
O
O
H3C
CH3
Ph
N
N
Aq. HBr, AcOH
HN
NH
CF3
Ph
reflux
(85%)
F3C
(S,R,R,S)-14
CF3
F3C
(R,R)-15
Tf2O, 2,6-lutidine
PhCF3, 0 °C
(88%)
O
Scheme 4. Trifluoromethylation of chiral bis-imine 11.
SO2CF3
N
HN
CF3
F3C
(R,R)-16
ORTEP of 16
For the converson of the unreacted diastereomer (S,S,S,S)-12 to
urea (S,S)-16, a two-step process was requried (Scheme 7). First,
cleavage of the N--phenethyl groups with Pd/C and hydrogen in
EtOH (or MeOH) provided good conversion to the TREC-
diamine 17.18 Then, reaction of 17 with phosgene, under less
forcing conditions than used previously, afforded the urea 15 in
79% yield.
Although the diastereomers 12 and 13 were not separable by
chromatography or crystallization, treatment with excess
phosgene (1.2–1.5 equiv) resulted in reaction only with the
(S,R,R,S)-13 isomer to give the urea (S,R,R,S)-14 (oil) and
recovered unreacted diamine 12 (Scheme 5). The two products
were easily separated by silica gel chromatography and upon
isolation of (S,S,S,S)-12 (low-melting solid, mp 81 °C)
crystallization occurred upon standing. The structure and the
absolute stereochemistry of 12 was determined by single-crystal
X-ray diffraction analysis and illustrated as an ORTEP in
Scheme 5.
Scheme 7. Synthesis of 2-oxazolidinone (S,S)-15.
To compare the chiral ligand DTFTI with the DPTI and DTBTI
from Figure 1, we prepared several mixed-ligand dirhodium
Scheme 5. Stereodivergent synthesis of urea 14.
complexes using the methods previously reported.1a Of the 14
O
H3C
CH3
Ph
possible Rh2(RCO2)n(DTFTI*4–n
)
complexes that can be
1. COCl2 / PhMe
cat. DMAP
(S,S,S,S)-12
+
prepared, we isolated nine distinct catalysts to test in the
asymmetric [2+1]-cycloaddition reaction. Out of these, the three
that provided the highest enantioselection are illustrated in Figure
2. It’s important to note that the stereochemical arrangement in
the (R,R)-16 DTFTI ligand is opposite to that from ligands (R,R)-
3 and (R,R)-4 due to the higher CIP priority of the CF3 group vs
phenyl. Thus, all the catalysts with ligands of the (R,R)-
configuration gave the same (S) enantiomer of the cyclopropene
product. However, opposite trends were observed with the
DTFTI complexes. For example, complexes 2 and 4, which are in
the M-cis-anti configuration (Figure 1), gave the highest %ee in
the Rh2(RCO2)2(L)2 series of catalysts, whereas with the DTFTI
ligand, complex 18 having the P-cis-anti configuration
performed the best.
N
N
Ph
2,6-lutidine
(S,R,R,S)-13
F3C
CH2Cl2, 23–50 °C
CF3
(1:1.2)
2. Chromatography
(S,R,R,S)-14
(74%)
+
H3C
Ph
CH3
Ph
NH HN
F3C
CF3
(S,S,S,S)-12
unreacted
ORTEP of 12
Reaction of pure 14 with HBr in refluxing acetic acid gave 2-
oxazolidinone (R,R)-15 (mp 223 °C, onset by DSC) in 85% yield
after 18 h (Scheme 6).17 The corresponding cleavage using 12 N
HCl at reflux was too slow to be useful. The conversion of 15 to
the mono-triflimide 16 required careful selection of conditions
because of solubility issues, the need to avoid bis-triflimide
formation and the unsuitability of many bases, but an excellent
process was developed as indicated in Scheme 6 which provided
Cyclopropene formation using the mixed-ligand complexes 18–
20 led to product of lower %ee at room temperature than
reactions at 0 °C. The P-cis-anti complex 18 containing one
trifluoroacete and one acetate bridging ligand, was superior to all
other complexes examined, resulting in 98% ee of the
cyclopropene product 1. In comparison, the mixed-ligand

4
Tetrahedron
complexes 2 and 4, with the M-cis-anti, both gave product of
91% ee at room temperature.
For the mixed-ligand complexes in the Rh2(RCO2)1(L)3 series,
complex 19 (Figure 2), which was derived from 18 (P-cis-anti)
by displacement of an acetate with DTFTI, gave 91% ee in the
[2+1]-cycloaddition reaction. In contrast, a similar trend was
observed with complex 5, which was derived from the complex
having the M-cis-anti configuration, and provided the highest
enantioselection the (1,3)-Rh2(RCO2)1(L)3 series.
Conclusion
In summary, we successfully developed
a useful bis-
trifluoromethylation process to prepare the chiral C2-symmetric
R,R-oxazolidinone 15 and its enantiomer from a common
intermediate. These compounds were then converted to the
corresponding N-triflimides and used for the synthesis of various
mixed-ligand dirhodium(II) complexes. Three such chiral mixed-
ligand Rh(II) complexes catalyzed the asymmetric [2+1]-
cycloaddition reaction of ethyl diazoacetate with terminal
acetylenes to form chiral cyclopropenes in high ee.
The mixed-ligand complex 20, having the configuration
corresponding to complex 5, (confirmed by X-ray structural
analysis, and both derived from their respective M-cis-anti
complexes), gave only 75% ee for the cyclopropene product. This
result in accordance with the trend reversal observed in the
DTFTI ligand series.
H
CO2Et
0.5 mol % 18–20
N
C5H11
(S)
= (R,R)-16
O
+
C5H11
H
CH2Cl2, 23 ˚C
CO2Et
N2
%ee, 1
N
O
N
O
O
O
Acknowledgments
O
Rh
O
O
Rh
N
N
O
Rh
N
N
Rh
O
N
Rh
O
N
Rh
O
We are grateful to Drs. Richard Staples and Robin Weatherhead
for the X-ray crystallographic data analyses, Shaw Huang, Bill
Collins and Mahender Reddy Karla for NMR spectral data
assistance at Harvard University, and to Pfizer Inc. for a research
grant.
O
O
O
O
O
19
81% ee
91% ee (0 ˚C)
20 (X-ray)
73% ee
75% ee (0 ˚C)
18, P-cis-anti
78% ee
95% ee (0 ˚C)
98% ee (0 ˚C)a
Supplementary Material
= OAc
O
a O
O
O
= OCOCF
3
Synthetic procedures and characterization data for the new
compounds and Rh2(II) complexes reported herein (PDF); X-ray
crystallographic data for compounds (±)-9, (S,S,S,S)-12, (R,R)-
16, and Rh2(OAc)(DTFTI)3 20 (CIF). This material is available
free of charge via the Internet.
ORTEP of 20
(2,1)-Rh2(OAc)(R,R-DTFTI)3
References
Figure 2. Enantioselective cyclopropenation with mixed-ligand
Rh2(II) catalysts 18–20 containing chiral ligand (R,R)-16.
6
For reviews on synthesis of vicinal diamines, see: (a) Corey, E. J. Pure
Appl. Chem. 1990, 62, 1209−1216; (b) Lucet, D.; La Gall, T.; Mioskowski, C.
Angew. Chem. Int. Ed. 1998, 37, 2580–2627; (c) Corey, E. J.; Kurti, L.
Enantioselective Chemical Synthesis; Direct Books Publishing and Elsevier,
2010; pp 18−23, 41-2; (d) Karlsson, S.; Lindberg, J.; Sorensen, H. Org.
Process Res. Dev., 2013, 17, 1552–1560.
1 (a) Lou, Y.; Remarchuk, T. P.; Corey, E. J. J. Am. Chem. Soc.
2005, 127, 14223−14230; (b) Lou, Y.; Horikawa, M.; Kloster, R. A.;
Hawryluk, N. A.; Corey, E. J. J. Am. Chem. Soc. 2004, 126,
8916−8918.
2 The descriptors for axial-chirality are M (counter clockwise) and P
(clockwise), whereas “cis” refers to two ligands adjacent to one
another and “anti” indicates the ligands are orientated on the
opposite rhodium.
7 H. J. Emeleus and R. N. Haszeldine, J. Chem. Soc., 1949, 2948–
2952.
8 (a) Trifluoromethylation of carbonyl compounds with the [CF3ZnI]
reagent was reported by: Kitazume, T.; Ishikawa, N. Chemistry
Letters, 1981, 1679–1680, and also employed within our group
(Letavic, M., 1993) to prepare 1-(2-bromophenyl)-2,2,2-
trifluoroethan-1-ol; (b) For a good review of nucleophilic
trifluoromethylation, see: Rubiales, G.; Alonso, C.; Martinez de
Margorta, E.; Palacios, F. ARKIVOC 2014 (ii) 362–405.
9 Bis-imine 8 is a crystalline white solid: (a) Neumann, W. L.; Rogic,
M.; Dunn, T. J. Tetrahedron Lett. 1991, 32, 5865–5868; (b) Liu, L.;
Ishida, N.; Ashida, S.; Murakami, M. Org. Lett. 2011, 13, 1666–
1669.
3 The experimental pKa value of each ligand was determined by
measuring the pH at one-half the equivalence point in a 1:1 EtOH-
H2O mixture: DTBTI (pKa = 13.2) and DPTI (pKa = 11.9).
4 (a) McClinton, M.A.; McClinton, D. A. Tetrahedron 1992, 48,
6555–6666; (b) Seebach, D. Angew. Chem. Int. Ed. Engl. 1990, 29,
1320–1367.
5
Zard and coworkers unexpectedly prepared a meso 1,2-CF3-
diamine derivative via a radical dimerization: (a) Gagosz, F.; Zard,
S. Org Lett., 2004, 5, 2655–2657; (b) Zard, S.; Gagosz, F. Fr.
Demande (2004) FR 2844264.
10 Alternative solvents such as THF and Et2O resulted in complex
reaction mixtures, as did warming the reaction mixtures in DMF.
Using a 5:1 CH3CN/DMF mixture resulted in only the mono-
addition product being formed.

5
11 For preparation of bis-imine (S,S)-11, see: (a) Alvaro, G.;
Grepioni, F.; Savoia, D. J. Org. Chem. 1997, 62, 4180; (b) Roland,
S.; Mangeney, P.; Alexakis, A. Synthesis 1999, 228–230; (c) Roland,
S.; Mangeney, P. Eur. J. Org. Chem. 2000, 611–616; (d) Alvaro, G.;
Grepioni, F.; Grilli, S.; Maini, L.; Martelli, G.; Savoia, D. Synthesis
2000, 581–587; (e) Sun, X.; Wang, S.; Sun, S.; Zhu, J.; Deng, J.
Synlett 2005, 2776−2780.
12 The mixture of diastereomers is isolated as yellow oil that
solidifies on standing to a wax-like product. Crystallization to
separate the diastereomers was unsuccessful.
13 For zinc activation, see: Organic Syntheses, Coll. Vol. 6, p.289
(1988); Vol. 53, p.86 (1973).
14 (a) Other approaches examined to access 12/13 include: Pd(0)-
catalyzed homocoupling of CF3-imidoyl iodide (Amii, H.; Kohda,
M.; Seo, M.; Uneyama, K. Chem. Commun., 2003, 1752–1753), and
(b) the widely used TMSCF3 reagent (Prakash, G. K. S.; Mandal, M.
J. Am. Chem. Soc. 2002, 124, 6538–6539) in the presence of TBAT.
Both approaches failed to give any product; (c) Preparation of the
[CF3Cd] reagent (Burton, D. J. Wiemers, D. M. J. Am. Chem. Soc.
1985, 107, 5014–5015), prepared from difluorodibromomethane
(CBr2F2, bp 24 °C) and Cd, resulted in trace mono CF3-addition
product from 11.
15 (a) Our group recently reported the preparation of 1,2-pentafluoro-
ethyl-1,2-diaminoethane where high (98:2) diastereoselectivity
starting with chiral bis-imine 11 was achieved. Vemula, R.; Wilde,
N. C.; Goreti, R.; Corey, E. J. Org. Lett., 2017, 19, 3883–3886; (b)
For discussion of the Cram chelate model, see: Bambridge, K.;
Begley, M. J.; Simpkins, N. S. Tetrahedron Lett. 1994, 35, 3391–
3394.
16
(a) A similar NMR study was done with CF3Br and Zn, see:
Francese, C.; Tordeux, M.; Wakselman, C. Tet. Lett, 1988, 29, 1029–
1030. The CF3Br is much less reactive with Zn than CF3I; (b)
Prakash, G. K. S.; Wang, F.; Zhang, Z.; Haiges, R.; Rahm, M.;
Christe, K. O.; Mathew, T.; Olah, G. A. Angew. Chem. Int. Ed.,
2014, 53, 11575–11578.
17
Hydrogenolysis of the N--phenethyl group with Pd/C and H2 (up
to 400 psi) and with acid additives (HCl or AcOH) was unsuccessful.
18 The freebase of 17 (generated by Et2O/saturated aqueous
NaHCO3) is volatile and not easily isolated in good yield (see SI).