Stereoinversion in the diastereoselective acylation of benzoxazine derivatives with 2-aryloxypropionyl chlorides

Article abstract of DOI:10.1007/s11172-019-2550-z

A comparative study of the kinetic resolution of racemic derivatives of 3,4-dihydro-3-methyl-2H-[1,4]benzoxazine using racemic 2-aryloxypropionyl chlorides was performed. It was found that the acylation of racemic amines with racemic 2-(1-naphthyloxy)propionyl chloride leads to amides enriched with (3R*,2′R*)-diastereomers, while the acylation with 2-phenoxypropionyl chloride gives predominantly (3R*,2′S*)-amides. Quantum chemical modeling of the process of kinetic resolution at the COSMO-CH₂Cl₂-B3LYP-D3-gCP/def2-TZVP//B3LYP-D3-gCP/def2-SVP level of theory was performed. The computational results are in a good agreement with the experimental data.

Full text of DOI:10.1007/s11172-019-2550-z

Russian Chemical Bulletin, International Edition, Vol. 68, No. 6, pp. 1257—1263, June, 2019
1257
Stereoinversion in the diastereoselective acylation of benzoxazine
derivatives with 2-aryloxypropionyl chlorides*
S. A. Vakarov,a M. A. Korolyova, D. A. Gruzdev, M. G. Pervova, G. L. Levit, and V. P. Krasnova
a
a,b
a
a,b
аI. Ya. Postovsky Institute of Organic Synthesis, Ural Branch of the Russian Academy of Sciences,
2
2/20 ul. S. Kovalevskoi, 620990 Ekaterinburg, Russian Federation.
Fax: +7 (343) 374 1189. E-mail: savakarov@ios.uran.ru
b
Institute of Chemical Engineering,
Ural Federal University named after the first President of Russia B. N. Yeltsin,
9 ul. Mira, 620002 Ekaterinburg, Russian Federation
1
A comparative study of the kinetic resolution of racemic derivatives of 3,4-dihydro-3-methyl-
Н-[1,4]benzoxazine using racemic 2-aryloxypropionyl chlorides was performed. It was found
2
that the acylation of racemic amines with racemic 2-(1-naphthyloxy)propionyl chloride leads
to amides enriched with (3R*,2´R*)-diastereomers, while the acylation with 2-phenoxypropio-
nyl chloride gives predominantly (3R*,2´S*)-amides. Quantum chemical modeling of the
process of kinetic resolution at the COSMO-CH Cl -B3LYP-D3-gCP/def2-TZVP//B3LYP-
2
2
D3-gCP/def2-SVP level of theory was performed. The computational results are in a good
agreement with the experimental data.
Key words: acylation, heterocyclic amines, acyl chlorides, 2-alkyloxy acids, stereoselectivity,
stereoinversion, density functional theory (DFT).
Kinetic resolution of racemates is one of the methods
for obtaining individual enantiomers of various classes of
We have systematically studied the stereoselective
reactions of racemic heterocyclic amines with 2-hydroxy
1
4—8
compounds, including practically valuable heterocyclic
acyl chlorides.
The COSMO-CH Cl -B3LYP-D3-
2 2
amines. This method is based on the difference in the rates
of the reactions of stereoisomers of a racemic substrate
with a chiral or achiral agent in the presence of a chiral
catalyst. It is relatively rare to observe a situation when
during kinetic resolution two structurally close chiral re-
solving agents of the same stereoconfiguration preferably
react with different enantiomers of the racemic substrate.
To refer to such a phenomenon, the term inversion of
gCP/def2-TZVP calculation method was demonstrated
to be applicable for establishing the reaction mechanism
and explanation of the reasons for the differences in the
7
kinetic resolution.
The purpose of the present work was to study the mu-
tual kinetic resolution of racemic 3-methylbenzoxazines
1a and 1b with racemic 2-(1-naphthyloxy)propionyl chlor-
ide (2b) structurally close to 2-phenoxypropionyl chloride
(2a), as well as to analyze the stereochemical results using
quantum chemical simulation of the diastereomeric tran-
2
stereoselectivity, or stereoinversion, is commonly used.
The synthetic precursors of many chiral catalysts,
chiral synthesis agents, and chiral derivatizing agents are
natural compounds such as carboxylic acids, alcohols, and
amines. These compounds are often available from natu-
ral sources only as a single stereoisomer; therefore, such
agents can be used to obtain enantiomerically pure amines
of one configuration, while the synthesis of its optical
antipode can be significantly more complex. Reactions in
which the inversion of stereoselectivity is observed are
under intense studies, since stereoinversion opens the way
to the synthesis of both stereoisomers of the racemic sub-
sition states (TSs) at the COSMO-CH Cl -B3LYP-D3-
2
2
gCP/def2-TZVP//B3LYP-D3-gCP/def2-SVP level. The
density functional theory (DFT) method, being not time
consuming and allowing one to accurately enough evalu-
ate noncovalent interactions, provided not only qualitative
but also quantitative agreement of the calculated and the
3
strate using the agents obtained from one chiral precursor.
*
Dedicated to Academician of the Russian Academy of Sciences
O. N. Chupakhin on the occasion of his 85th birthday.
1: X = H (a), F (b)
Published in Russian in Izvestiya Akademii Nauk. Seriya Khimicheskaya, No. 6, pp. 1257—1263, June, 2019.
066-5285/19/6806-1257 © 2019 Springer Science+Business Media, Inc.
1

1258
Russ. Chem. Bull., Int. Ed., Vol. 68, No. 6, June, 2019
Vakarov et al.
Scheme 1
2
: Ar = Ph (a), 1-naphthyl (b )
Compound
X
H
H
Ar
Ph
Compound
4a
4b
X
F
F
Ar
Ph
3
3
a
b
1-naphthyl
1-naphthyl
Reagents and conditions: 2a,b (0.5 equiv.), CH Cl , 6 h, –20 C.
2
2
experimental results of the acylation of heterocyclic amines
(3S,2´R)- and (3S,2´S)-diastereomers of amides 3b and
4b (the ratio of diastereomers ~1 : 1) starting from the
equimolar amounts of (S)-amines 1a and 1b and racemic
acyl chloride 2b in the presence of N,N-diethylaniline (as
an HCl acceptor), which were isolated from the reaction
mixture by column flash chromatography and character-
ized by a combination of physicochemical methods of
analysis.
7
with 2-aryloxy acyl chlorides. Chiral heterocycles 1a,b were
used by us earlier as model substrates for studying the ki-
netic resolution of racemic amines by carboxylic acid de-
rivatives, that is why we used them in the present work, too.
The selectivity factor s is the most important charac-
teristic of kinetic resolution, which is the ratio of reaction
rates of rapidly and slowly reacting enantiomers. To de-
termine the s value of the acylation of heterocycles 1a,b,
we used an approach based o n the reaction of racemic
reagents, i.e., the limiting version of mutual kinetic reso-
To assign the configuration of the predominant amides
3b and 4b, we acylated enantiopure (S)-amines 1a and 1b
with racemic acyl chloride 2b in the presence of N,N-di-
ethylaniline (as an HCl acceptor) at a ratio of
(S)-amine : acyl chloride : N,N-diethylaniline equal to
1 : 2 : 1. In this case, a kinetic resolution of the racemic acyl-
ating agent took place. The reaction gave diastereomeri-
cally enriched amides 3b and 4b and unreacted enantiomeri-
cally enriched acyl chloride 2b, which was hydrolyzed with
subsequent isolation of the corresponding acid. A com-
parison of the sign of optical rotation of the acid obtained
1
lution. In this case, the s value can be determined quite
accurately.9
The acylation of racemic amines 1a and 1b with acyl
7
chloride 2a was described in the work. The acylation of
amines (RS)-1a and (RS)-1b with its analog 2b was carried
out under the same conditions, namely, in dichlorometh-
ane at –20 C for 6 h at a 2 : 1 molar ratio of amine : acyl
chloride (Scheme 1).
1
0
The reaction led to a mixture of (3R,2´R)/(3S,2´S)-
and (3R,2´S)/(3S,2´R)-amides 3a,b and 4a,b (or (3R*,2´R*)-
and (3R*,2´S*)-amides, respectively). The ratio of dia-
stereomers (dr) of amides in the reaction mixtures equal
to the parameter s was determined by GLC (Table 1).
To unambiguously assign the peaks of diastereomers
in the GLC chromatograms, we synthesized mixtures of
with the literature data showed that the unreacted acyl
chloride belongs to the (R)-series and, therefore, its opti-
cal antipode reacted with (S)-3-methylbenzoxazines faster,
with (3R*,2´R*)-diastereomers predominating among the
amides obtained by the reaction of racemic reagents.
Earlier, we showed that the reaction of racemic amines 1a
and 1b with racemic acyl chloride 2a predominantly gave
7
(
3S*,2´R*)-amides. This result means that a stereoinver-
sion occurred on going from agent 2a (Ar = Ph) to acyl
chloride 2b (Ar = 1-naphthyl), with the selectivity factor
in the case of acylation of amines 1a and 1b with acyl
chloride 2b being significantly lower (s 7.8 and 17, respec-
tively ) than in the case of agent 2a (s 25 and 62, respec-
tively) (see Table 1). At the same time, difluoro-contain-
ing 3-methylbenzoxazine 1b reacted with acyl chloride 2b
much more selectively than its non-fluorinated analog 1a
Table 1. Stereochemical results of acylation of amines 1a
and 1b with acyl chlorides 2a and 2b
Amine
Acyl
chloride
Dr of amides
Selectivity
factor s
*
*
*
* a
(R ,S )/(R ,R )
1
1
a
b
2a
96.2 : 3.8
11.4 : 88.6
98.4 : 1.6
5.5 : 94.5
25b
2
b
17.8
b
2a
62
(
s 17 and 7.8, respectively).
2
b
17
7
Recently, we showed for the first time that the
a
COSMO-CH Cl -B3LYP-D3-gCP/def2-TZVP//B3LYP-
An average value of the results of two experiments accord-
2 2
ing to GLC.
D3-gCP/def2-SVP method is applicable for the study of
kinetic resolution in diastereoselective acylation. It was
b
The results are taken from the work.7

Stereoinversion in acylation of benzoxazines
Russ. Chem. Bull., Int. Ed., Vol. 68, No. 6, June, 2019
1259
7
found that the acylation of chiral benzoxazines with
-hydroxycarboxylic acyl chlorides follows the S 2-like
of heterocyclic amines 1a,b with acyl chloride 2a. The
2
DFT calculation of the Gibbs free energy of activation of
two competing reactions as the difference between the
Gibbs free energies of TS and the corresponding PC (see
Ref. 15) or stable intermediates allows one to achieve
a good agreement between the calculated and the experi-
mental values.
N
mechanism through the transition state corresponding to
the energy maximum in the reaction pathway. According
to the proposed model, the addition of the nucleophile to
the carbonyl carbon atom of the acylating agent occurs
1
6
7
with the simultaneous elimination of the chloride ion.
To explain the observed stereoinversion, as well as the
differences in the stereoselectivity of acylation of benz-
oxazines 1a and 1b with acyl chloride 2b, we performed
DFT calculations of the parameters of the reaction of
Earlier, we similarly calculated the Gibbs free energies
of activation for the reactions of (R)-amines 1a,b with acyl
7
chloride 2a (Table 2).
It was found that the Gibbs free energies of activation
of the reactions leading to (3R,2´S)-amides 3a and 4a are
lower than the activation energies of the reactions with
the formation of (3R,2´R)-amides, which is consistent
with the experimental data. The simulation of the acylation
of amines 1a,b with acyl chloride 2b showed that the ac-
tivation energies of the reactions leading to (3R,2´S)-
amides 3a and 4a, on the contrary, are higher than the
activation energies of the competing reactions giving
(3R,2´R)-diastereomers as the products (Fig. 1, see Table 2).
The quantum chemical computational results are consis-
tent with the observed stereoinversion of the acylation of
amines 1a,b on going from 2-phenoxy-substituted acyl
chloride 2a to 2-(1-naphthyloxy)-substituted agent 2b.
For example, the difference in the Gibbs activation energies
for the pairs of reagents (S)-2b/(R)-1a and (R)-2b/(R)-1a
(
R)-amines 1a and 1b with (RS)-2-(1-naphthyloxy)pro-
pionyl chloride (2b) at the COSMO-CH Cl -B3LYP-D3-
2
2
gCP/def2-TZVP or def2-SVP)//B3LYP-D3-gCP/def2-
SVP level and compared the results with the computational
7
data for the reaction of (R)-amines 1a and 1b and acyl
chloride 2a.
Before the quantum chemical modeling of the acyla-
tion of amines 1a and 1b with acyl chloride 2b, we carried
out a conformational search and selected the most favor-
able conformations of the aliphatic ring of amines, in which
the methyl group occupies an axial position.
The selectivity factor can be calculated from the dif-
ference in the Gibbs free energies of activation of the
reactions of enantiomers of a racemic substrate with
a chiral agent or a catalyst.11 In a qualitative evaluation
of stereoselectivity, one often compare not the activation
energies, but the Gibbs free energies (G) of diastereomeric
–
1
is 3.17 kJ mol , with the experimental value being
4.31 kJ mol–
1
.
1
1—13
TSs.
This simplified approach is based on the as-
Analysis of the TS geometries showed that in the case
of the reactions of amines 1a and 1b with both acyl chlor-
ide 2a and acyl chloride 2b, there is a "sandwich" π-stacking
of the aromatic fragments of reagents in both (3R,2´S)-TS,
which is not observed in the case of (3R,2´R)-TS (Fig. 2).
At the same time, the Gibbs absolute free energies of
formation for (3R,2´S)-TS 1a/2a and 1b/2a are lower than
for (3R,2´S)-TS 1a/2b and 1b/2b. It can be assumed that the
π-stacking of aryl fragments of reagents in the case of acyl-
ation of amines with reagent 2a, stabilizing (3R,2´S)-TS,
is more favorable than the aromatic interactions in
sumption that the Gibbs free energies of pre-reaction
complexes (PCs) do not differ or the differences in them
are negligibly small.12 Its application is justified in the
situations when the PC is unstable (the equilibrium is
shifted towards the reagents), as well as in the cases when
1
3
the structure of the reagent complexes is unclear or it is
necessary to single out the predominant mechanism from
1
4
the many possible pathways of the reaction of one PC.
Using the BLYP/TZVP methods, we showed the for-
mation of stable diastereomeric PCs during the acylation
Table 2. Calculated values of the Gibbs free activation energies (ΔG , kJ mol–1)
#
at –20 C in CH Cl for the reaction of (R)-amines 1a and 1b with (R)- and
2
2
(
S)-acyl chlorides 2a,b at the COSMO-CH Cl -B3LYP-D3-gCP/def2-TZVP//
B3LYP-D3-gCP/def2-SVP level
2
2
Reagents
ΔG#
ΔΔG#
Calculation Experimenta
(
(
(
(
S)-2a/(R)-1a, (R)-2a/(R)-1a
S)-2a/(R)-1b, (R)-2a/(R)-1b
S)-2b/(R)-1a, (R)-2b/(R)-1a
28.19, 34.82
38.66, 45.45
43.21, 40.04
45.74, 41.04
6.63
6.79
3.17
4.70
6.77b
c
b
8.68
4.31
5.94
c
S)-2b/(R)-1b, (R)-2b/(R)-1b
a
b
c
#
#
The experimental ΔΔG values are calculated by the formula ΔΔG = –RT ln s.
7
The results are taken from the work.
Geometric optimization of the complex of reagents was carried out from the
model structure of TS.

1260
Russ. Chem. Bull., Int. Ed., Vol. 68, No. 6, June, 2019
Vakarov et al.

G253/kJ mol–1
G253/kJ mol–1
110
a
b
1
10
00
1
100
90
80
70
60
50
40
30
20
10
0
(
R,S)-TS
1.08
9
8
7
6
5
4
3
2
1
0
0
0
0
0
0
0
0
0
0
8
(
R,S)-TS
1.72
7

G#253(R,R)
40.04

G#253(R,S)
3.21
(R,R)-PS
4
41.04
(
R,S)-PC
8.51
2
Reagents
Reagents
(R,S)-Amide
(R,R)-Amide
r
r
Fig. 1. The calculated coordinate of the reaction (r) of amine 1a and acyl chloride 2b leading to (R,S)- (a) and (R,R)-amides (b) in
dichloromethane at –20 С (COSMO-CH Cl -B3LYP-D3-gCP/def2-TZVP//B3LYP-D3-gCP /def2-SVP).
2
2
(
3R,2´S)-TS 1a/2b and 1b/2b. According to the quantum
chemical calculations, the benzene ring of amines 1а and
b in (3R,2´S)-TS 1a/2b and 1b/2b is located above the
CBS basis set is 2.76 times higher in the second case than
in the first.
1
According to the DFT calculation, amine 1b is less
nucleophilic than its analogue 1a. This leads to the fact
that the nitrogen atom of the amino group of benzoxazine
1b reacts with the carbonyl carbon atom of acyl chlorides
2a and 2b at a shorter distance compared with the reactions
of the same acyl chlorides and amine 1a. The shortening
of the N—C bond in TS probably increases the steric re-
quirements to the molecules of reagents in the process of
stereodifferentiation, and, as a consequence, the selectiv-
ity factor also increases.
central double bond of the naphthyl fragment of acyl
chloride 2b. According to the literature data,17 such an
arrangement is characteristic of the benzene—naphthalene
"
sandwich" dimeric π—π-complexes. The calculated val-
ues of the bonding energy of benzene molecules in the
sandwich" dimer1 and the formation energy of the ben-
8
"
1
7
zene—naphthalene complex are known. The formation
energy of the π—π-complex calculated by the standard
method of coupled clusters with a complete CCSD(T)/
(
3R,2´S)-TS
(3R,2´R)-PC
Fig. 2. The geometry of diastereomeric transition states of the reaction of 1a with 2b calculated at the COSMO-CH Cl -B3LYP-D3-
2
2
gCP/def2-TZVP//B3LYP-D3-gCP/def2-SVP level.

Stereoinversion in acylation of benzoxazines
Russ. Chem. Bull., Int. Ed., Vol. 68, No. 6, June, 2019
1261
gramming at 10 С min–1 to 280 С (30 min plateau). The injec-
In conclusion, we carried out a comparative study of
the mutual kinetic resolution of heterocyclic amines and
racemic 2-aryloxypropionyl chlorides. It was established
that the acylation of racemic 3,4-dihydro-3-methyl-
tor and detector temperatures were 250 and 300 C, respec-
tively, nitrogen was a carrier gas, the flow separation was 1 : 30,
–
1
the flow through the column was 1.0 mL min with the injection
of 1.0 μL of amide solution in MeCN with a concentration of
2
H-[1,4]benzoxazines by racemic 2-(1-naphthyloxy)-
–
1
1
—3 mg mL
.
propionyl chloride leads to amides with a predominance
of (3R*,2´R*)-diastereomers, in contrast to the reaction
of these amines with 2-phenoxypropionyl chloride, in
which (3R*,2´S*)-amides are predominantly formed.
Quantum chemical modeling of diastereoselective acyla-
tion was performed. It was established that the calculated
Gibbs free energies of activation of the reactions of
(
RS)-2-(1-Naphthyloxy)propionic acid. A solution of α-naphth-
ol (3.98 g, 27.6 mmol) in THF (30 mL) was added dropwise
to a cooled to 0 C suspension of NaH (1.04 g, 43.3 mmol) in
THF (30 mL) with vigorous stirring. The resulting mixture was
stirred for another 15 min at the same temperature, followed by
the addition of a solution of methyl 2-bromopropionate (9.79 g,
5
5
8.6 mmol) in THF (30 mL). Then, the mixture was stirred for
h at 20 C, concentrated in vacuo to 20 mL, poured into water
(
R)-amines and racemic 2-(1-naphthyloxy)propionyl
(
60 mL), and extracted with chloroform (4×15 mL). The or-
chloride agree with the data obtained experimentally. It
was shown that the chosen calculation method is sensitive
to a slight change in the structure of the agents, since it
leads to the results reflecting the observed in experiments
inversion of stereoselectivity in the acylation of 3,4-di-
hydro-3-methyl-2H-[1,4]benzoxazines when a phenoxy
group in the structure of the acylating agent is replaced
with the (1-naphthyloxy) group.
ganic layer was washed with water (3×20 mL), dried with Na SO ,
2
4
and concentrated in vacuo. The residue was dissolved in EtOH
(
100 mL), the resulting solution was cooled to 0 C, and 2 M
NaOH (27.6 mL) was added dropwise with stirring. The reaction
was stirred for 24 h at 20 C and concentrated in vacuo, the
residue was acidified with concentrated HCl and extracted with
chloroform (4×25 ml). The organic layer was washed with brine
(
3×20 mL), dried with Na SO , and concentrated in vacuo. The
2 4
residue, a brown oil crystallizing on standing, was recrystallized
from a mixture of hexane—EtOAc (12 : 1) to obtain (RS)-2-(1-
naphthyloxy)propionic acid. The yield was 3.82 g (64%),
a colorless crystalline powder, m.p. 154—156 C (hexane—EtOAc)
(cf. Ref. 9: m.p. 154—155 C). Found (%): C, 72.50; H, 5.39.
Experimental
(
RS)-3,4-Dihydro-3-methyl-2H-[1,4]benzoxazine (1a),19
RS)-3,4-dihydro-3-methyl-7,8-difluoro-2H-[1,4]benzoxazine
(
(
[
C13H12O . Calculated (%): C, 72.21; H, 5.59. HPLC: τ 20.2
3
(
R
)
1
9
1
13
1b),
(3S)-3,4-dihydro-3-methyl-2H-[1,4]benzoxazine
min; τ(S) 39.7 min. The H and C NMR spectra are identical
20
10
(S)-1a], and (3S)-3,4-dihydro-3-methyl-7,8-difluoro-2H-[1,4]-
to those published earlier.
2
1
benzoxazine [(S)-1b] were synthesized by known procedures.
(RS)-2-(1-Naphthyloxy)propionyl chloride (2b). Oxalyl chlor-
ide (1.13 mL, 12.9 mmol) and DMF (5 μL) were added to a solu-
tion of (RS)-2-(1-naphthyloxy)propionic acid (1.00 g, 4.62 mmol)
in anhydrous dichloromethane (50 mL). The reaction mixture
was stirred for 6 h at ~20 C and concentrated in vacuo. The yield
*
*
*
*
Acyl chloride 2a, amides (3R ,2´S )-3a, (3R ,2´S )-4a, as well
as mixtures of diastereomeric (3R,2´S)- and (3S,2´S)-amides 3a
4
and 4a were described earlier. Other reagents were commer-
cially available. Solvents were purified using standard procedures.
1
13
19
1
H, C, and F NMR spectra were recorded on a Bruker
was 1.08 g (quantitative). H NMR, δ: 1.86 (d, 3 H, Me, J = 6.8 Hz);
Avance 500 spectrometer (500, 126, and 470 МHz, respectively),
5.12 (q, 1 H, 2-CH, J = 6.8 Hz); 6.68—6.70 (m, 1 H, Ar);
7.29—7.37 (m, 1 H, Ar); 7.47—7.54 (m, 3 H, Ar); 7.77—7.84
(m, 1 H, Ar); 8.28—8.34 (m, 1 H, Ar).
1
using tetramethylsilane as an internal standard. H NMR spec-
trum of acyl chloride 2b was recorded on a Bruker Avance 400
1
spectrometer (400 МHz). H NMR spectra of amides 3a and 3b
Stereoselective acylation of racemic amines 1a,b with acyl
chlorides 2a,b (general procedure). A solution of the correspond-
ing acyl chloride (0.5 mmol) in dichloromethane (5 mL) was
added to a solution of аmine 1a or 1b (1.0 mmol) in the same
solvent (5 mL) at –20 C and the mixture was thermostated for
6 h at –20 C. Then, the reaction mixture was sequentially washed
with 4 М HCl (2×4 mL), brine (4×5 mL), 5% aqueous NaHCO3
were recorded in DMSO-d at 100 C, spectra of other com-
6
pound, in CDCl at ~20 C. Melting points were determined on
3
a Stuart SMP3 apparatus (Barloworld Scientific, UK). Elemental
analysis was performed on a Perkin—Elmer 2400 II automatic
CHNS-O analyzer. Analytical TLC was performed on Sorbfil
plates (Imid Ltd., Russia), silica gel (230—400 mesh) (Alfa Aesar,
UK) was used for flash chromatography. Optical rotation of
scalemic 2-(1-naphthyloxy)propionic acid were measured on
a Perkin—Elmer M341 polarimeter (Perkin—Elmer Instruments,
(2×5 mL), and H O (2×5 mL), dried with Na SO , and con-
2
2
4
centrated. The residue was analyzed by GLC. The diastereomers
of amides were isolated by recrystallization.
–
1
–1
USA). Specific rotation is expressed in deg mL g dm , con-
(3R*,2´R*)-3,4-Dihydro-3-methyl-N-[2´-(1ʺ-naphthyloxy)-
centration of solutions, in g (100 mL)– . HPLC analysis of
1
*
*
propionyl]-2H-[1,4]benzoxazine [(3R ,2´R )-3b]. The yield was
76.4 mg (44%) after recrystallization from a mixture of hexane—
EtOAc (1 : 1), a white amorphous powder. Found (%): C, 76.33;
2
-(1-naphthyloxy)propionic acid was performed on a Shimadzu
LC-20 Prominence instrument: a Chiralcel OD-H column
(
250×4.6 mm, Daicel Corp., Japan), detection at 220 nm, the
H, 6.20; N, 4.19. C H NO . Calculated (%): C, 76.06; H, 6.09;
2
2
21
3
–
1
eluent flow rate was 1 mL min , the mobile phase hexane—
N, 4.03. GLC: τ(
33.9 min; τ(R*,R*)-3b 32.9 min; (R*,S*)—
R*,S*)-3b
i
1
Pr OH—CF COOH (60 : 0.8 : 0.2 ). GLC analysis of amides 3b
(R*,R*) (10.3 : 89.7). H NMR, δ: 1.10 (d, 3 H, Me of benz-
oxazine, J = 6.8 Hz); 1.69 (d, 3 H, MeCH, J = 6.3 Hz); 3.97
3
and 4b was carried out using a Shimadzu GC 2010 gas chromato-
graph with a flame ionization detector, a ZB-5 quartz capillary
column (length 30 m, diameter 0.25 mm, film thickness 0.25 μm);
the initial column temperature was 40 C (3 min plateau), pro-
(dd, 1 H, 2-CHB of benzoxazine, J = 11.0 Hz, J = 2.8 Hz);
1
2
4.14 (dd, 1 H, 2-CH of benzoxazine, J = 11.0 Hz, J = 1.6 Hz);
A
1
2
4.73 (qdd, 1 H, 3-CH of benzoxazine, J = 6.8 Hz, J = 2.8 Hz,
1
2

1262
Russ. Chem. Bull., Int. Ed., Vol. 68, No. 6, June, 2019
Vakarov et al.
J3 = 1.6 Hz); 5.61 (q, 1 H, 2-CH, J = 6.3 Hz); 6.70—6.77 (m, 1 H,
7.78—7.87 (m, 1 H, Ar); 8.14—8.20 (m, 0.5 H, Ar); 8.20—8.25
(m, 0.5 H, Ar).
Ar); 6.82—6.89 (m, 2 H, Ar); 7.00—7.07 (m, 1 H, Ar); 7.26—7.32
(
m, 1 H, Ar); 7.42—7.53 (m, 3 H, Ar); 7.62—7.68 (m, 1 H, Ar);
7,8-Difluoro-3,4-dihydro-3-methyl-N-[2´-(1ʺ-naphthyloxy)-
propionyl]-2H-[1,4]benzoxazine [4b (mixture of diastereomers)].
The yield was 153 mg (80%) after flash chromatography (eluent
hexane—EtOAc (95 : 5), a white amorphous powder. Found (%):
1
3
7
1
1
1
.78—7.85 (m, 1 H, Ar); 8.14—8.20 (m, 1 H, Ar). C NMR, δ:
5.14, 17.22, 45.62, 69.25, 70.79, 105.60, 116.45, 119.94, 120.60,
21.64, 122.85, 124.80, 125.06, 125.34, 125.71 (2 C), 126.51,
27.35, 134.15, 145.69, 152.21, 168.54.
C, 68.98; H, 5.09; F, 9.81; N, 3.63. C22H19F NO . Calculat-
2
3
*
*
(
3R ,2´R )-7,8-Difluoro-3,4-dihydro-3-methyl-N-[2´-(1ʺ-
ed (%): C, 68.92; H, 5.00; F, 9.71; N, 3.65. GLC: τ(
33.9
R,S)-4b
*
*
1
naphthyloxy)propionyl]-2H-[1,4]benzoxazine [(3R ,2´R )-4b].
The yield was 95.8 mg (50%) after recrystallization from a mix-
ture of hexane—EtOAc (1.7 : 1), a white crystalline powder. M.p.
min; τ(S,S)-4b 32.8 min; (R,S)—(S,S) (54.4 : 45.6). H NMR, δ:
1.06 (d, 1.8 H, Me of benzoxazine (R,S), J = 6.5 Hz); 1.11
(d, 1.2 H, Me of benzoxazine (S,S), J = 6.8 Hz); 1.657 (d, 1.8 H,
CHMe (R,S), J = 6.3 Hz); 1.662 (d, 1.2 H, CHMe (S,S),
J = 6.3 Hz); 4.06 (dd, 0.4 H, 2-CHB of benzoxazine (S,S),
1
44—147 C (hexane—EtOAc). Found (%): C, 69.02; H, 5.08;
F, 9.68; N, 3.49. C22H19F NO . Calculated (%): C, 68.92;
2
3
H, 5.00; F, 9.71; N, 3.65. GLC: τ
33.9 min;
J = 11.1 Hz, J = 2.7 Hz); 4.20 (dd, 0.6 H, 2-CH of benzox-
(
R*,S*)-4b
1 2 B
*
*
*
*
1
τ(
R*,R*)-4b
32.8 min; (R ,S )—(R ,R ) (2.6 : 97.4). H NMR, δ:
azine (R,S), J = 11.2 Hz, J = 2.6 Hz); 4.31 (dd, 0.4 H, 2-CH
1 2 A
1
.11 (d, 3 H, Me of benzoxazine, J = 6.8 Hz); 1.66 (d, 3 H,
CHMe, J = 6.3 Hz); 4.06 (dd, 1 H, 2-CHB of benzoxazine,
J = 11.1 Hz, J = 2.7 Hz); 4.31 (dd, 1 H, 2-CH of benzoxazine,
of benzoxazine (S,S), J = 11.1 Hz, J = 1.4 Hz); 4.35 (dd, 0.6 H,
1 2
2-CHA of benzoxazine (R,S), J = 11.0 Hz, J = 1.5 Hz); 4.77
1
2
(qdd, 0.4 H, 3-CH of benzoxazine (S,S), J = 6.8 Hz, J = 2.8 Hz,
1
2
A
1
2
J = 11.1 Hz, J = 1.4 Hz); 4.77 (qdd, 1 H, 3-CH of benzoxazine,
J = 6.8 Hz, J = 2.8 Hz, J = 1.5 Hz); 5.65 (q, 1 H, 2-CHMe,
J3 = 1.5 Hz); 4.91 (qdd, 0.6 H, 3-CH of benzoxazine (R,S),
J1 = 6.8 Hz, J2 = 2.9 Hz, J3 = 1.5 Hz); 5.57 (q, 0.6 H,
2-CHMe (R,S), J = 6.6 Hz); 5.65 (q, 0.4 H, 2-CHMe (S,S),
J = 6.3 Hz); 6.74—6.95 (m, 2 H, Ar); 7.30—7.36 (m, 0.4 H, Ar of
acid (S,S)); 7.37—7.41 (m, 0.6 H, Ar (R,S)); 7.43—7.57 (m, 4 H,
Ar); 7.80—7.85 (m, 1 H, Ar of acid); 8.10—8.17 (m, 0.4 H, Ar of
1
2
1
2
3
J = 6.3 Hz); 6.74—6.87 (m, 2 H, Ar of acid, H(6) of benzoxazine);
.30—7.36 (m, 1 H, Ar of acid); 7.43—7.57 (m, 4 H, Ar of acid,
H(5) of benzoxazine); 7.80—7.85 (m, 1 H, Ar Acyl); 8.10—8.17
7
1
3
(
m, 1 H, Ar Acyl). C NMR, δ: 15.27, 16.84, 45.59, 69.81, 72.11,
05.72, 106.91 (d, J = 18.0 Hz); 119.42 (dd, J1 = 7.8 Hz,
J2 = 3.9 Hz); 120.66, 120.88, 121.51, 125.04, 125.32, 125.77,
1
9
1
acid (S,S)); 8.18—8.23 (m, 0.6 H, Ar of acid (R,S)). F NMR,
δ: 2.08 (ddd, 0.6 F, C(8)F (R,S), J = 21.0 Hz, J = 8.1 Hz,
1
2
1
1
1
26.50, 127.36, 134.13, 136.09 (dd, J = 9.9 Hz, J = 3.0 Hz);
J3 = 2.1 Hz); 2.20 (ddd, 0.4 F, C(8)F (S,S), J1 = 21.0 Hz,
1
2
38.85 (dd, J = 243.8 Hz, J = 15.4 Hz); 146.85 (m); 152.10,
J = 7.8 Hz, J = 2.0 Hz); 21.07—21.27 (m, 1 F, C(7)F).
1
2
2
3
68.73. 19F NMR, δ: 2.12—2.29 (m, 1 F, C(8)F); 21.07—21.26
Kinetic resolution of acyl chloride 2b by (S)-аmines 1a and 1b
(
m, 1 F, C(7)F).
(general procedure). A solution of acyl chloride 2b (141 mg,
Synthesis of mixtures of diastereomeric amides 3b and 4b
general procedure). A solution of (RS)-2-(1-naphthyloxy)pro-
0.6 mmol) in toluene (2 mL) was added to a solution of (S)-аmine
(
1a or 1b (0.3 mmol) and PhNEt (44.8 mg, 0.3 mmol) in toluene
2
pionyl chloride (117 mg, 0.5 mmol) in CH Cl (5 mL) was
(4 mL) at 20 C. The mixture was thermostated for 24 h at 20 C
and then concentrated to dryness in vacuo. The residue was dis-
2
2
added to a solution of (S)-аmine 1a or 1b (0.5 mmol) and PhNEt2
75 mg, 0.5 mmol) in CH Cl (10 mL) at 20 C. The mixture
(
solved in MeCN (10 mL) and a saturated solution of Na CO3
2
2
2
was thermostated for 24 h at 20 C and then sequentially washed
(10 mL) was added. The reaction mixture was vigorously stirred
for 1 h, MeCN was evaporated in vacuo. The aqueous solution
was extracted with CHCl3 (2×5 mL). The organic layer was
sequentially washed with water (2×5 mL), 4 M HCl (2×4 mL),
brine (4×5 mL), and water (2×5 mL). The organic layer was
with 4 M HCl (2×5 mL), brine (3×15 mL), 5% aq. NaHCO3
(
10 mL), and water (2×15 mL). The organic layer was separated,
dried with Na SO , and concentrated. The residue was purified
2
4
by flash chromatography (eluent hexane—EtOAc).
,4-Dihydro-3-methyl-N-[2´-(1ʺ-naphthyloxy)propionyl]-
H-[1,4]benzoxazine [3b (mixture of diastereomers)]. The yield
3
dried with Na SO and concentrated in vacuo. The residue was
2 4
2
analyzed by GLC.
was 158 mg (91%) after flash chromatography (eluent hexane—
EtOAc (95 : 5)), a white amorphous powder. Found (%): C, 76.04;
The alkaline aqueous solutions (after extraction with CHCl3)
were combined, acidified with 4 M HCl to pH 1—2, and ex-
H, 6.27; N, 4.33. C H NO . Calculated (%): C, 76.06; H, 6.09;
tracted with CHCl (2×5 mL). The organic layer was washed
2
2
21
3
3
N, 4.03. GLC: τ
(
(
34.0 min; τ(S,S)-3b 33.0 min; (R,S)—(S,S)
with brine (2×5 mL), dried with Na SO , and concentrated
2 4
(
R,S)-3b
1
45.0 : 55.0). H NMR, δ: 1.05 (d, 1.5 H, Me of benzoxazine
R,S), J = 6.8 Hz); 1.10 (d, 3 H, Me of benzoxazine (S,S),
in vacuo. The residue was purified by flash chromatography
(eluent hexane—EtOAc) and analyzed by HPLC on a chiral
stationary phase.
J = 6.8 Hz); 1.63 (d, 1.5 H, MeCH (R,S), J = 6.5 Hz); 1.69
d, 3 H, MeCH (S,S), J = 6.3 Hz); 3.97 (dd, 0.5 H, 2-CH of benz-
(
(R)-2-(1-Naphthyloxy)propionic acid was obtained by ki-
netic resolution of (RS)-2b with (S)-аmine 1b. The yield was
46.7 mg (72%), a white crystalline powder. M.p. 124—125 C
B
oxazine (S,S), J = 11.0 Hz, J = 2.8 Hz); 4.12—4.17 (m, 1 H,
1
2
2
4
-CH of benzoxazine (R,S) and 2-CH of benzoxazine (S,S));
B
A
2
0
.20 (dd, 0.5 H, 2-CHA of benzoxazine (R,S), J = 11.0 Hz,
(hexane—EtOAc), [α]
+31.4 (c 0.54, CHCl ) (cf. Ref. 10:
3
1
D
3
2
0
J2 = 1.6 Hz); 4.73 (qdd, 0.5 H, 3-CH of benzoxazine (S,S),
[α]D +58.9 (c 1.08, CHCl )). Found (%): C, 72.02; H, 5.79.
J = 6.8 Hz, J = 2.8 Hz, J = 1.6 Hz); 4.85 (qdd, 0.5 H, 3-CH
of benzoxazine (R,S), J = 6.8 Hz, J = 2.8 Hz, J = 1.7 Hz);
C13H12O . Calculated (%): C, 72.21; H, 5.59. Ee 60%, HPLC:
1
2
3
3
1
13
τ(R) 20.2 min; τ(S) 39.7 min. The H and C NMR spectra are
1
2
3
1
0
5
.58 (q, 1 H, 2-CHMe (R,S), J = 6.5 Hz); 5.61 (q, 1 H, 2-CHMe
identical to those published earlier.
(
(
S,S), J = 6.3 Hz); 6.70—6.77 (m, 0.5 H, Ar (S,S)); 6.80—6.88
m, 2 H, Ar); 6.92—6.96 (m, 0.5 H, Ar (R,S)); 7.00—7.07 (m, 1 H,
Quantum chemical calculations. Graphic modeling and pri-
mary optimization of transition state geometries was carried out
using molecular mechanics and molecular dynamics methods
Ar); 7.26—7.32 (m, 0.5 H, Ar (S,S)); 7.37—7.42 (m, 0.5 H, Ar
R,S)); 7.42—7.53 (m, 3 H, Ar); 7.62—7.69 (m, 1 H, Ar),
2
2,23
(
(Ammp software package) and the VEGA ZZ program.
As

Stereoinversion in acylation of benzoxazines
Russ. Chem. Bull., Int. Ed., Vol. 68, No. 6, June, 2019
1263
a result, model structures of all diastereomers of transition states
were obtained.
6. S. A. Vakarov, D. A. Gruzdev, L. Sh. Sadretdinova, M. I.
Kodess, A. A. Tumashov, E. B. Gorbunov, G. L. Levit, V. P.
Krasnov, Chem. Heterocycl. Compd., 2018, 54, 437.
7. M. A. Korolyova, S. A. Vakarov, D. A. Gruzdev, G. L. Levit,
V. P. Krasnov, Eur. J. Org. Chem., 2018, 4577.
DFT calculations based on the model structures were per-
formed using the ORCA 3.0.3.2
4,25
In each EFSP calculation,
geometrical optimization was performed using D3 type dispersion
2
6,27
corrections.
The solvent (CH Cl ) was taken into account
8. S. A. Vakarov, D. A. Gruzdev, E. N. Chulakov, G. L. Levit,
V. P. Krasnov, Russ. Chem. Bull., 2019, 68, 481.
9. J. Brandt, C. Jochum, I. Ugi, P. Jochum, Tetrahedron, 1977,
33, 1353.
10. L. Giampietro, A. Ammazzalorso, I. Bruno, S. Carradori,
B. De Filippis, M. Fantacuzzi, A. Giancristofaro, C. Mac-
callini, R. Amoroso, Chem. Biol. Drug. Des., 2016, 87, 467.
11. H. B. Kagan, J. C. Fiaud, Top. Stereochem., 1988, 18,
249—330.
2
2
2
8
using the conductor‐like screening model (COSMO). The
geometry and energy of the starting compounds, transition states,
and then pre-reaction complexes (at a fixed CO—Cl bond length
of 1.82 Å) were determined using the Beke—Lee—Yang—Parr
2
9,30
hybrid meta-GGA functional (B3LYP)
and Ahlrichs´s basic
sets def2-SVP and def2-TZVP.31
The search for transition states at the first stage and the
initial calculation of the Hessian in the gas phase were carried
out at the B3LYP-D3/def2-SVP level, changing the length of
the key N—CO bond. Each iteration during geometry optimiza-
tion was accompanied by determining the energy of single-point
calculations as a sum of the total energy and the D3 dispersion
correction together with the geometric counterpoise correction
12. B. Wanner, I. Kreituss, O. Gutierrez, M. C. Kozlowski, J. W.
Bode, J. Am. Chem. Soc., 2015, 137, 11491.
13. S. M. Bakalova, F. J. S. Duarte, M. K. Georgieva, E. J.
Cabrita, A. G. Santos, Chem. Eur. J., 2009, 15, 7665.
14. S. E. Allen, S.-Y. Hsieh, O. Gutierrez, J. W. Bode, M. C.
Kozlowski, J. Am. Chem. Soc., 2014, 136, 11783.
15. E. Larionov, M. Mahesh, A. C. Spivey, W. Yin, H. J. Zipse,
J. Am. Chem. Soc., 2012, 134, 9390.
3
2
gCP for the basis superposition error. The calculations were
speeded up using a RIJCOSX approximation.2
5,33
The total
thermal energies were calculated as a sum of the total electron
energy, zero-point vibration energy, and thermal corrections
16. N. Carter, L. Xiabing, L. Reavey, A. J. H. M. Meijer,
I. Coldham, Chem. Sci., 2018, 9, 1352.
(
vibrational, rotational, translation). In calculation of the Gibbs
free energy, the entropy component value was chosen to corre-
spond to the symmetry number n = 1 (since all the reagents are
chiral compounds).34
17. W. Wang, T. Sun, Y. Zhang, Y.-B. Wang, J. Chem. Phys.,
2015, 143, 114312.
18. S. Tsuzuki, K. Honda, T. Uchimaru, M. Mikami, K. Tanabe,
J. Am. Chem. Soc., 2002, 124, 104.
The authors are grateful to M. I. Kodess, M. A. Ezhi-
kova, and E. G. Matochkina for recording the NMR
spectra, L. Sh. Sadretdinova for carrying out HPLC ana-
lyzes, S. V. Sharf for assistance in carrying out quantum
chemical calculations.
Equipment of the Centre for Joint Use "Spectroscopy
and Analysis of Organic Compounds" at the I. Ya. Pos-
tovsky Institute of Organic Synthesis of the Ural Branch
of the Russian Academy of Sciences was used in the work.
Quantum chemical calculations were performed on the
URAN supercomputer at the N. N. Krasovsky Institute
of Mathematics and Mechanics of the Ural Branch of the
Russian Academy of Sciences.
19. EU Pat. 0047005; Chem. Abstrs, 1982, 97, 55821b.
20. D. A. Gruzdev, G. L. Levit, V. P. Krasnov, E. N. Chulakov,
L. Sh. Sadretdinova, A. N. Grishakov, M. A. Ezhikova, M. I.
Kodess, V. N. Charushin, Tetrahedron: Asymmetry, 2010,
2
1, 936.
1. P. A. Slepukhin, D. A. Gruzdev, E. N. Chulakov, G. L. Levit,
V. P. Krasnov, V. N. Charushin, Russ. Chem. Bull., 2011,
0, 955.
2. A. Pedretti, L. Villa, G. Vistoli, J. Comput.-Aided Mol. Des.,
004, 18, 167.
2
6
2
2
23. A. Pedretti, L. Villa, G. Vistoli, J. Mol. Graph., 2002, 21, 47.
24. F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2, 73.
2
2
2
2
5. F. Neese, F. Wennmohs, A. Hansen, U. Becker, Chem. Phys.,
009, 356, 98.
6. S. Grimme, S. Ehrlich, L. Goerigk, J. Comput. Chem., 2011,
2, 1456.
7. S. Grimme, A. Hansen, J. G. Brandenburg, C. Bannwarth,
Chem. Rev., 2016, 116, 5105.
8. S. Sinnecker, A. Rajendran, A. Klamt, M. Diedenhofen,
F. Neese, J. Phys. Chem. A., 2006, 110, 2235.
9. A. D. Becke, J. Chem. Phys., 1993, 98, 1372.
0. A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
2
This work was financially supported by the Russian
Foundation for Basic Research (Project No. 18-33-00027
mol_а).
3
References
2
3
1
. M. R. Maddani, J.-C. Fiaud, H. B. Kagan, in Separation of
31. F. Weigend, R. Ahlrichs, J. Phys. Chem. Chem. Phys., 2005,
Enantiomers: Synthetic Methods, Ed. M. Todd, Wiley-VCH,
Weinheim, 2014, p. 13–74.
7
, 3297.
3
3
3
2. H. Kruse, S. Grimme, J. Chem. Phys., 2012, 136, 154101.
2
3
4
. M. Bartók, Chem. Rev., 2010, 110, 1663.
3. F. Neese, J. Comput. Chem., 2003, 24, 1740.
. T. Tanaka, M. Hayashi, Synthesis, 2008, 21, 3361.
. S. A. Vakarov, D. A. Gruzdev, L. Sh. Sadretdinova, E. N.
Chulakov, M. G. Pervova, M. A. Ezhikova, M. I. Kodess,
G. L. Levit, V. P. Krasnov, Tetrahedron: Asymmetry, 2015,
4. A. Fernández-Ramos, B. A. Ellingson, R. Maena-Paneda,
J. M. C. Marques, D. G. Truhlar, Theor. Chem. Acc., 2007,
1
18, 813.
2
6, 312.
5
. S. A. Vakarov, D. A. Gruzdev, E. N. Chulakov, L. Sh.
Sadretdinova, A. A. Tumashov, M. G. Pervova, M. A. Ezhi-
kova, M. I. Kodess, G. L. Levit, V. P. Krasnov, V. N. Cha-
rushin, Tetrahedron: Asymmetry, 2016, 27, 1231.
Received February 6, 2019;
in revised form April 19, 2019;
accepted April 25, 2019