Slight structural modulation around a pivotal bond: high impact on enantiomeric stability

Article abstract of DOI:10.1039/d1nj03178c

Based on aN-arylthiazoline scaffold, 22 structures with a N-C_arylchiral axis were synthesized exhibiting a huge molecular diversity for the four flanking substituents around the pivotal bond. The determination of their corresponding rotational

Full text of DOI:10.1039/d1nj03178c

NJC
View Article Online
View Journal
PAPER
Slight structural modulation around a pivotal
bond: high impact on enantiomeric stability†
Cite this: DOI: 10.1039/d1nj03178c
Daniel Farran, * Nicolas Vanthuyne,
Giulia Bossa, Vincent Belot,
Muriel Albalat, Marion Jean and Christian Roussel
Based on a N-arylthiazoline scaffold, 22 structures with a N–Caryl chiral axis were synthesized exhibiting a
huge molecular diversity for the four flanking substituents around the pivotal bond. The determination of
their corresponding rotational barriers by thermal kinetic experiments or from the analysis of plateau shape
chiral HPLC chromatogram allowed to rank these compounds according to their enantiomeric stability:
4 rotamers, 5 isolable atropisomers (i.e. not enough robust to be handled without risk of racemization) and
13 stable atropisomers. The influence of the flanking substituents was investigated showing that a minor
structural modification may result in a drastic change on the value of the rotational barrier. All these results
offer a set data on structure-rotational barrier relationships very helpful to design molecules exhibiting chiral
axis or to optimize the enantiomeric stability of such structure. To complete this study, the racemization
pathways were examined thanks to X-ray analysis and DFT calculations, highlighting the importance of the
aromatic ring distortion during the transition state on the energetic cost of the rotation.
Received 30th June 2021,
Accepted 30th July 2021
DOI: 10.1039/d1nj03178c
rsc.li/njc
of atropisomerism was revealed as an exciting issue in drug
discovery with promising opportunities.
Introduction
According IUPAC, atropisomers are defined as ‘‘a subclass of
The dynamic aspect of atropisomerism has to be addressed
conformers that can be isolated as separate chemical species with high care during the design and the optimization of
and which arise from restricted rotation about a single bond’’.1 molecules such as drugs or catalysts. The Oki’s precept6 is
For instance, atropisomers are observed in the presence of a often recalled to fix the border between rotamers (fast rotation)
chiral axis i.e. ‘‘an axis about which a set of ligands is held so and atropisomers (restricted rotation): it states that atropi-
that it results in a spatial arrangement that is not superposable somerism concerns isolable isomers with a half-life time of
on its mirror image’’ and lead consequently to a pair of interconversion of at least 1000 seconds corresponding to a
enantiomers. During the last decades, chemists took advan- rotational barrier of 92.8 kJ molꢀ1 at 25 1C. This statement is
tages of this type of chirality to develop molecules exhibiting convenient but not entirely satisfactory mainly because not
very high potential,2 especially in homogenous asymmetric enough accurate regarding the feature ‘‘isolable’’ of a pure
catalysis.3 However, since this stereoisomerism depends on a enantiomer. Laplante et al. proposed a practical approach to
dynamic process of interconversion between both enantiomers, tackle this issue: stereoisomers possessing a chiral axis are
such structures were not considered as appropriate for a classified into three groups depending on their rotational
pharmaceutical purpose. Very recently, this dogma was chal- barriers.5 Class I consists of rotamers that spin fast about
lenged by medicinal chemists due to the need to explore a chiral axis from one conformer to the other. Class III includes
larger chemical space for drug candidates.4,5 Thus, the concept enantiomeric stable atropisomers which have to be seen simi-
larly to unracemizable enantiomers. Finally, atropisomers with
Aix Marseille Univ., CNRS, Centrale Marseille, iSm2, Marseille, France.
intermediate enantiomeric stability are comprised in class II;
these compounds can be isolated but under specific conditions
(such as low temperature) which makes impossible their use for
a given application. Herein, we classified compounds with
E-mail: daniel.farran@univ-amu.fr
† Electronic supplementary information (ESI) available: Copies of 1H NMR, 13C
NMR and 19F NMR spectra for all new compounds. Kinetic experiments and
determination of rotational barriers (with solvent and temperature). Calculated
rotational barriers obtained with B3LYP/6-311G(3d,3p), M06-2X/6-311G(3d,3p) chiral axis according a similar ranking, taking in mind that
and PBE0/def2-TZVPP methods. X-Ray data for compound (Sa)-1c, (Sa)-2c (Sa)-3c
the boundary between two consecutive groups should be envi-
(Sa)-6c. ECD and UV spectra for compounds exhibiting a DGarot 4 100 kJ molꢀ1
.
saged as slightly flexible in order to adjust to the intended
purpose of the molecule. For instance, a catalyst with a chiral
axis operating in a reaction at room temperature can be
considered as a stable atropisomer when its rotational barrier
Selected geometrical parameters (angle, dihedral angle, bond length) obtained
from PBE0/def2-TZVPP method for GS and both TS of all compounds. CCDC
1955658–1955661. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/d1nj03178c
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem.

View Article Online
Paper
NJC
providing compounds belonging to each of the three classes
of stereoisomers mentioned previously.10 Moreover, the rota-
tional barrier is weakly dependent of the temperature as
demonstrated by the small and constant entropy parameters.8
Thus, this homogenous series can be considered as reliable to
compare rotational barriers even if they are not evaluated
exactly at the same temperature. Besides these crucial points,
it is worth noting that a huge molecular diversity is easily
accessible on N-arylthiazolines since they are synthesized from
commercially available ortho-substituted anilines and a-halogeno
ketones.
Fig. 1 Racemization in N-arylthiazoline series.
Results and discussion
is greater than 115 kJ molꢀ1 (i.e. a half-life time of 91 days at
25 1C).
The ortho halogenated N-arylthiazoline-2-thiones 1 represent a
relevant starting point since a wide range of values of rotational
barriers is covered from the fluoro derivative 1a to the iodo
compound 1d (Table 1). A plateau shape was observed during
liquid chromatography on a chiral stationary phase for com-
pound 1a and the corresponding barrier to rotation was esti-
mated at 82.5 kJ molꢀ1 according to the Trapp and Schurig
equation,11 implying that 1a belongs to the class II. The three
other structures are included in class III due to the higher steric
demand of their respective halogen atoms compared to the
fluorine.
Density functional theory (DFT) calculations were performed
for series 1 using three different functional/basis sets: B3LYP/
6-311G(3d,3p), M06-2X/6-311G(3d,3p) and PBE0/def2-TZVPP.12
Calculations provided the barriers and allowed to explore the
energies associated to the two possible pathways for the rota-
tion about the pivotal N–Caryl bond (Fig. 1). During the rotation,
the N-aryl ortho halogen is passing in front of the methylene
group (Fig. 1: pathway A) or in front of the sulphur atom (Fig. 1:
pathway B). Whatever the method employed, calculated bar-
riers are in line with experimental ones, PBE0/def2-TZVPP
reproducing the experimental data most closely (Table 1).
Previous experimental data from the steric scale based on a
N-arylthiazoline-2-thione scaffold gave a methyl group bigger
The aim of this paper is to highlight that a minor structural
modification introduced close to a chiral axis generates major
outcomes on the enantiomeric stability of an atropisomer.
Although many rotational barriers are reported in literature,7
few studies have been carried out on series of compounds in
order to evaluate the impact of molecular diversity on DGraot
values.8,9 The set of results reported hereafter provides some
general trends on the structure-rotational barrier relationship.
These data are of significant importance in the design and the
structure optimization for compounds of interest: structural
modifications can be envisaged to modulate the rotational
barrier either by increasing it to access stable atropisomers or
by decreasing it to lead to rotamers developed as single flexible
compounds.
To achieve our goal, a N-arylthiazoline template (Fig. 1) with
a restricted rotation about the N–Caryl bond appears extremely
appropriate. In these series, the resulting rotational barriers
represent a proper view of the steric bulk of substituents borne
on the blocking positions around the pivotal bond, whereas the
contribution of electronic aspects is negligible.8 In addition,
the barriers to rotation on this scaffold are very sensitive to
steric variations giving rise to a large range of DGarot and
Table 1 Rotational barriers of ortho halogenated N-arylthiazoline-2-thionesab
DGacalc
DGacalc
Via Ac
DGacalc
Via Ac
X
1
DGarot
Via Ac
Via Bc
2
DGarot
Via Bc
3
DGarot
Via Bc
F
1a
1b
1c
1d
82.5
133.3
145.5
155.5
83.7
137.8
152.3
167.4
112.9
154.1
163.1
171.6
2a
2b
2c
2d
113.7
159.2
165.9
168.1
113.7
163.7
175.2
189.5
143.1
175.2
181.0
189.4
3a
3b
3c
3d
114.7
161.6
167.3
168.6
117.2
167.5
178.4
191.6
138.2
173.8
179.9
187.4
Cl
Br
I
a
b
c
Rotational barriers are reported in kJ molꢀ1
.
Solvent and temperature are given in ESI. Calculated rotational barriers through pathway A or B
using PBE0 functional and def2-TZVPP basis set.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
connected to the angle between the endocyclic nitrogen atom
and the substituent borne in position 4 of the thiazoline ring.
This angle was measured at 120.41 for 4-methyl thiazoline-2-
thione 3c and 122.31 for the six-membered ring derivative 2c,
compared to 131.81 found for the five-membered ring homo-
logue 1c leading to a much easier rotation around the pivotal
bond in this latter case.
DFT-calculated data reproduced properly the experimental
DGarot for these two new series 2 and 3 (Table 1), although DGacalc
for iodo derivatives 2d and 3d are overestimated whatever the
level of theory used.12 Nevertheless, despite this deviation, it is
Fig. 2 X-Ray structures of the (Sa) enantiomer of bromo derivatives.
than a chlorine which allowed to hypothesize none Cl/S elec- important to mention that the general evolution of DGcaalc
trostatic repulsion in the TS and thus suggested that the values in both series, from the fluoro derivative to the iodo
racemization proceeds through the transition state 1 (TS1) in derivative, matches suitably to experimental results given the
series 1.8 This assumption was supported by the calculations wide range of the corresponding rotational barriers.14 As
achieved herein which gave the pathway A as the most favour- noticed for series 1, calculations indicated that the racemiza-
able with a lower energetic cost.
tion mechanism proceed via TS1, except for compounds 2d and
From these preliminary results, an expanded study to the 3c for which any assumptions may not be stated due to
whole four flanking substituents was carried out, and we first insignificant energy gap between pathways A vs. B, and also
concentrated our efforts on the cyclopentene borne by the for 3d, for which the computational study suggests the opposite
thiazoline core. Two modifications were envisioned on this racemization mechanism via TS2.15 It is particularly interesting
moiety of the skeleton: the extension to a six-membered carbo- to focus on this latter aspect since it helps to elucidate an
cycle (series 2) and the removal of the carbocycle accompanied experimental result which can be considered as odd at the first
by the installation of a methyl in position 4 of the thiazoline sight: DGarot of 3d (168.6 kJ molꢀ1 in 1,2,4-trichlorobenzene at
ring (series 3).13 The syntheses of the desired N-(o-substituted- 214 1C) similar to DGarot of 3c (167.3 kJ molꢀ1 in 1,2,
phenyl)-thiazoline-2-thiones were achieved using a protocol 4-trichlorobenzene at 214 1C), i.e. a difference of only 1.3 kJ molꢀ1
previously described:8 treatment of ortho-halogenated aniline despite the higher steric demand of an iodine vs. a bromine. A
with carbon disulfide to generate the dithiocarbamate salt plot of enantiomerization barriers of series 1 vs. series 3 allows to
which was allowed to react with 2-chlorocyclohexanone or detect a peculiar behaviour for compound 3d which accounts
chloroacetone to furnish 2 or 3 respectively. Preparative HPLC for the poor correlation (Fig. 3; r2 = 0.9702 for the whole data vs.
on chiral support on the so-called (S,S) Whelk-O1 column r2 = 0.9928 without iodine derivatives; red solid line represents
provided both enantiomers and rotational barriers were then linear regression for F, Cl and Br derivatives). In this context, the
determined by kinetic study of the thermal racemisation of comparison of the rotational barrier of 3d to the three others
atropisomers at a given temperature. The kinetic studies were compounds of series 3 requires special care since two different
carried out in similar solvents (chlorobenzene, 1,2-dichloro- racemization pathways are presumably involved. The same
benzene or 1,2,4-trichlorobenzene) to eliminate misinterpreta- trend seems to emerge when comparing series 1 to series 2
tion caused by a hypothetical solvent effect. The resulting Gibbs (Fig. 3; r2 = 0.9797 for the whole data vs. r2 = 0.9957 without
free energies of activation DGarot were obtained from the Eyring iodine derivatives; blue solid line represents linear regression for
equation and are gathered in Table 1.
F, Cl and Br derivatives), but the phenomenon is here less
For a given halogen atom in ortho position, compounds 2 and 3 significant and does not allowed to clearly rule for a reverse
possess very similar rotational barriers (Table 1). We obtained a DGarot racemization pathway for 2d versus 2a–c.
value for compound 2b of 159.2 kJ molꢀ1 in 1,2-dichlorobenzene at
180 1C vs. 161.6 kJ molꢀ1 for 3b under the same conditions. These
very close results revealed that substituents in position 4 on the
thiazoline ring of 2 and 3, i.e. methylene of a six-membered ring and
methyl respectively, exhibited a similar spatial requirement. Moreover,
these two structural modifications increased notably the rotational
barriers compared to homologues 1. A significant example concerns
compounds 1a and 2a that only differ by one atom on the size of their
carbocycles. This minor modulation generated a noteworthy increase
of 31.2 kJ molꢀ1 between the DGraot of 1a and 2a, and allowed moving
from the class I/class II border to the class II/class III one.
Careful examinations of the structures obtained by X-ray
crystal analyses of the three N-(o-bromophenyl)-thiazoline-2-
thiones 1c, 2c and 3c help to rationalize experimental DGraot
values (Fig. 2). The resulting rotational barriers are directly Fig. 3 Experimental rotational barriers for series 1 vs. series 2 and 3.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem.

View Article Online
Paper
NJC
A peripheral unexpected aspect raised by the plotting of g are characteristic of the aromatic pattern. Thus both
depicted on Fig. 3 concerns the amplitude on the DGarot value moieties of the N-arylthiazoline-2-thiones behave indepen-
involved by identical structural modifications carried out on dently of each other: the TS geometries of the thiazolinethione
different series. Indeed, the rotational barriers for series 1 ring are governed by the five membered ring/six membered
exhibited a greater sensitivity to the size of the ortho substituent ring/methyl substituents, while the nature of the halogen atom
than the six-membered ring and 4-Me derivatives, 2 and 3 series controls the TS geometries of the aromatic ring.
respectively, as showed by the slope of the linear regression
which is 41 when comparing the five-membered ring fluoro, g are similar in both transition states (e.g. for 3b: 127.41 in TS1
chloro and bromo derivatives with their homologues. vs. 127.21 in TS2), whereas pathway A lead to higher values for
For a given N-arylthiazoline-2-thione, the values of angle
To gain understanding on the conformational parameters angles a and b than pathway B (e.g. angle a/angle b for 2b:
involved in the two possible rotational pathways, information 125.61/132.61 in TS1 vs. 123.11/129.91 in TS2). In contrast, the
on molecular geometries can be gleaned from the computa- torsion of the aromatic ring is greater for TS2 than TS1 (e.g.
tional study. The values for angles between the bonds bearing a C10–C60–C40–C30 dihedral angle c for 2b: 4.81 in TS1 vs. 7.41 in
flanking substituent and the bond connected to the chiral axis TS2) resulting in a higher energy penalty in the former case.
(Fig. 4: angles a, b and g), as well as the torsion of the aromatic Taking into account that the computational study argues in
ring (dihedral angle c: C10–C60–C40–C30 in Fig. 4), the dihedral favour of a racemization mechanism via pathway A for com-
angle between the thiazoline ring and the aromatic ring (dihe- pounds 1a–d, 2a–c and 3a–b, the energetic cost for the twist of
dral angle f: C2–N3–C10–C60 in Fig. 4) and the pivotal bond the aromatic ring appears significant for the racemization
length C10–N3 are gathered in the ESI,† both for ground and process. Interestingly, the torsion of the aromatic ring is similar
transition states. These crucial parameters for the mechanism in TS1/TS2 for 2d (C10–C60–C40–C30 dihedral angle c: 8.11 in
of the rotation revealed that series 1, 2 and 3 exhibit the same TS1 vs. 8.51 in TS2), which could explain that none hypothesis
behaviour which is described hereafter.
on the racemization pathway may be established from calcula-
The optimized structures for the ground state (GS) showed tions in this case.
that the aryl ring was twisted with respect to the heterocyclic
In order to go further in the structure-rotational barrier
ring with C2–N3–C10–C60 dihedral angles f from 671 to 861. The relationship, we then investigated two related compounds to
geometry of the thiazoline moiety is identical within a series, previous series. Compound 4d displays two hydrogen atoms in
displaying none impact of the halogen atom. On the aromatic position 4 and 5 of the thiazoline ring (Table 2). Despite the
ring, a same variation for angle g is noticed in all series (from presence of an iodine atom in ortho-aryl position, the corres-
1191 for fluoro derivatives to 1211 for iodo derivatives) which ponding HPLC profile provided a single peak on several chiral
can be assigned to the increased steric bulk of the halogen stationary phases at 10 1C,16 meaning a rotational barrier lower
atom. It is also noteworthy that all these angles as well as than 80 kJ molꢀ1. Compound 4d unambiguously belongs to
C10–N3 bond lengths are in perfect agreement to those revealed class I, while the three iodo derivatives 1d, 2d and 3d are clearly
by crystal structures.
ranked in class III. This demonstrates the minimal require-
As expected, the transition states are almost planar. Com- ment of three substituents (different from hydrogen atoms) on
pared to GS, none significant stretch is observed for the pivotal the four flanking positions to access to a configurationally
bond N3–C10; on the other hand, angles a, b and g are wider, stable thiazoline-2-thione scaffold. Calculation using the
accompanied by torsions of thiazoline and aromatic rings to appropriate PBE0/def2-TZVPP method furnished a DGacalc of
allow the rotation around the chiral axis avoiding clashes 77.3 kJ molꢀ1 for 4d, with a quasi-planar transition state
(Fig. 4).
involving I/H5 and Harom/S interactions as expected
Within a series, the angles a and b are constant for a given (pathway A).
racemization pathway, without any influence of the halogen
atom borne by the aromatic ring: the values of a and b are
characteristic of the thiazoline pattern. Moreover, the angles g
are analogous for all the structures bearing the same halogen
regardless the series and the racemization pathway: the values
Table 2 Rotational barriers of ortho halogenated N-arylthiazoline-2-
thiones 4d and 5aab
DGacalc
Via Ac
77.3
DGacalc
Via Ac
152.9
DGarot
Via Bc
DGarot
Via Bc
o80
149.8
142.1
153.1
a
b
Rotational barriers are reported in kJ molꢀ1
.
Solvent and tempera-
c
ture are given in ESI. Calculated rotational barriers through pathway A
or B using PBE0 functional and def2-TZVPP basis set.
Fig. 4 Atom numbering in N-arylthiazoline and GS/TS for 3b.
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Comparison of rotational barriers between compounds 3a
Paper
Table 3 Rotational barriers of ortho halogenated N-arylthiazoline-2-
selenones 6 and N-arylthiazoline-2-ones 7ab
(114.7 kJ molꢀ1 in 1,2-dichlorobenzene at 62 1C) and 5a
(142.1 kJ molꢀ1 in 1,2-dichlorobenzene at 180 1C) offers another
set of attractive results. In these two structures, the four
substituents around the pivotal bond are identical, but they
are not arranged in the same way: the methyl is linked to the C4
of the thiazoline ring in 3a whereas it is borne in ortho-aryl
position for 5a. In both cases, the corresponding compounds
belong to class III. However, this minor modulation on the
skeletons induced an impressive gap of 27.4 kJ molꢀ1 on the
DGarot value. An additional consequence to this slight structural
change was revealed during chiral HPLC analysis: the two
enantiomers of 3a were well resolved on the (S,S) Whelk-O1
column (a = 1.49 and Rs = 6.65, 25 1C, heptane/isopropanol/
dichloromethane 70/20/10, 1 mL minꢀ1) whereas none enantio
discrimination was noticed for 5a in the same conditions,
illustrating a perturbation of the interaction between the chiral
stationary phase and the analyte in this last case.17
DGacalc
Via Ac
DGacalc
Via Ac
X
6
DGarot
Via Bc
7
DGarot
Via Bc
F
6a
6b
6c
6d
85.1
134.8
146.4
155.2
86.1
140.5
154.9
170.5
118.4
163.5
173.4
184.2
7a
7b
7c
7d
o80
56.1
102.6
116.6
133.6
78.2
109.3
116.3
123.0
Cl
Br
I
103.2
109.7
112.0
Rotational barriers are reported in kJ molꢀ1
.
Solvent and tempera-
a
b
c
ture are given in ESI. Calculated rotational barriers through pathway
A or B using PBE0 functional and def2-TZVPP basis set.
The DFT-calculated DGcaalc for 3a and 5a are in accordance
with the experimental results. However, although the computa- difference of merely 0.9 kJ molꢀ1. This similar behaviour is
tional study supports unambiguously pathway A for the race- consistent with a comparable steric bulk between sulphur and
mization process of 3a, calculated rotational barriers via TS1 selenium (van der Waals radii: S = 180 pm and Se = 190 pm).21
and TS2 are too close to plead for one of the two pathways in In addition, X-ray crystal analysis of N-(o-bromophenyl)-
the case of compound 5a. Moreover, the geometries of both thiazoline-2-selenone 6c displays an angle of 1311 between
transition states of 5a highlighted an important torsion of the the endocyclic nitrogen atom and the methylene substituent
aromatic ring (C10–C60–C40–C30 dihedral angle c for 5a: 11.71 in borne in position 4 of the thiazoline ring, i.e. exactly the same
TS1 and 10.31 in TS2, to be compared to 1.91 in TS1 for 3a) value observed for N-(o-bromophenyl)-thiazoline-2-thione 1c
which explains the extra energy required for the rotation of 5a (Fig. 2).
(Fig. 5).
In this context, the computational study furnished obviously
Studies were finally directed towards the nature of the exocyclic very close optimized GS and TS geometries for compounds of
chalcogen (Table 3). In order to investigate the impact of such series 6 compared to those observed in series 1, and the whole
modification, ortho halogenated N-arylthiazolineselenones 6 and remarks already mentioned above may be transposed.
N-arylthiazolinones 7 were synthesized.18 The parent thiazoli-
In contrast, the substitution of a thiocarbonyl group (series 1)
nethiones 1 were converted into their corresponding thiazolium by a carbonyl group (series 7) involved a significant impact on
salts by addition of methyl iodide. This intermediate salt allowed the value of barriers to rotation: a drastic decrease was
to access to the two targeted series: treatment with sodium observed, from 30 to 44 kJ molꢀ1 depending on the nature of
hydrogen selenide (obtained by reduction of selenium with the halogen.22 As a consequence, isolation of both enantiomers
sodium borohydride) furnished the desired selenones 619 whereas of the fluoro derivative 7a was not possible because of quick
a subsequent reaction with sodium methylate provided the oxy- interconversion.23 Indeed, the calculated rotational barrier was
genated derivatives 7.10,20
estimated at 56 kJ molꢀ1, switching thus this compound into
The corresponding barriers to rotation were measured as class I with respect to 1a ranked into the intermediate class II.
described above and revealed very close values when comparing Contrary to all others series, the chloro and bromo structures
results from sulphur series 1 to selenium series 6. For example, 7b and 7c are part of class II, while the N-(o-iodophenyl)-
compounds bearing a bromine atom in ortho-aryl position thiazoline-2-one 7d is located on the border class II/class III.
exhibited DGarot of 145.5 kJ molꢀ1 and 146.4 kJ molꢀ1 (both in
The DFT calculations revealed some notable aspects specific
1,2-dichlorobenzene at 180 1C) for 1c and 6c respectively, i.e. a to the transition states of the oxo compounds 7. Firstly, the
values for the angle b are smaller of 4.71 on average by
comparing the TS geometries of analogous structures from
series 7 to series 1; this can be attributed to the shorter CQO
bond (120 pm for CQO vs. 165 pm for CQS) generating less
flexibility.24 In addition, the TS geometries in series 7 are
characterized by less distortion on the aromatic moiety
with respect to the halogenated homologues in series 1 (e.g.
C10–C60–C40–C30 dihedral angle c for 7d: 5.51 in TS1 and 6.11 in
TS2, to be compared to 8.11 in TS1 and 8.01 in TS2 for 1d),
meaning a rotation process requiring less energy. Last but not
Fig. 5 Significant distortion on the aromatic moiety in both TS for 5a vs.
quasi-planar aromatic ring in the most favourable TS for 3a.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem.

View Article Online
Paper
NJC
least, contrary to all other series reported in this paper, the two Optical rotations were measured with
a
sodium lamp
transition states for the racemization of a given compound of (589 nm) and a double-jacketed 10 cm cell at 25 1C. The chiral
the series 7 exhibit very close geometries on the thiazoline HPLC analyses were performed on Agilent 1260 Infinity unit
moiety as illustrating by the values of angle a (e.g. in TS1/TS2: (pump G1311B, autosampler G1329B, DAD G1315D). The ana-
135.71/134.51 for 7d, to be compared to 135.91/131.21 for 1d), as lytical column (250 ꢁ 4.6 mm) used is (S,S)-Whelk-O1 from
well as for angle b (e.g. in TS1/TS2: 129.01/127.91 for 7d, to be Regis Technologies (Morton Grove, USA), except for 5a (Daicel
compared to 134.51/130.71 for 1d). Consequently, although Chiralpak IE). Retention times Rt in minutes, retention factors
DGacalc militate in favour of pathway A in the cases of fluoro ki = (Rti ꢀ Rt0)/Rt0 and enantioselectivity factor a = k2/k1 and
and chloro derivatives 7a and 7b, calculations do not discrimi- resolution Rs = 1.18(Rt2 ꢀ Rt1)/(w1 + w2) are given. Preparative
nate the two rotation pathways A and B for 7c, and support chiral separations were done with (S,S)-Whelk-O1 from Regis
pathway B for 7d.25
Technologies (Morton Grove, USA) with a mixture of heptane/
ethanol/chloroform (7/2/1) as mobile phase, except for 5a
(Daicel Chiralpak IE with heptane/ethanol, 9/1).
Conclusion
General synthesis of N-aryl-thiazoline-2-thione 2, 3, 4d and 5a
We reported the experimental and calculated rotational barriers of
22 N-arylthiazolines exhibiting a large molecular diversity on the
substituents located in the flanking positions of the chiral axis.
Indeed, several types of structural modifications were considered
on this same scaffold: substituting one group for another, swap-
ping the position of flanking substituents around the chiral axis,
or removing a group. The resulting rotational barriers (from DGrot
Distilled triethylamine (40 mmol) was added dropwise under
nitrogen atmosphere to a solution of an ortho-substituted ani-
line (20 mmol) in carbon disulfide (38 mL). The mixture was
stirred 24–48 h at r.t. Then the precipitate was filtered, washed
with Et2O and dried to give the dithiocarbamate salt. This salt
was used without any further purification and immediately
solubilised in acetonitrile (31 mL). 20 mmol of a-halogeno
ketone/aldehyde (2-chlorocyclohexanone for series 2, chloroa-
cetone for series 3, chloroacetaldehyde for 4d and 5a) were then
added dropwise at r.t. under nitrogen atmosphere. The mixture
was stirred 24 h at r.t. Then a 37% HCl solution (5 mL) was
added dropwise and the mixture was heated at reflux (oil bath)
for 20 minutes. The solvent was evaporated under reduced
pressure and water was added (50 mL). The mixture was
extracted with dichloromethane (3 ꢁ 50 mL), the organic layer
was dried on MgSO4 and evaporated under reduced pressure.
The desired product was purified by flash chromatography
(petroleum ether–dichloromethane, 100/0 - 0/100). 1a–d, 2a
and 3a are known compounds and their 1H NMR spectra are in
accordance with literature data.8,26
o 80 kJ molꢀ1 to DGrot = 169 kJ molꢀ1, i.e. DDGrot 4 89 kJ molꢀ1
)
allowed to accurately evaluate the impact of such modifications
on enantiomeric stability. To go further in understanding the
parameters involved, the two pathways of rotation around
the pivotal bond were studied and discriminated by exploring
the geometries of GS/TS with the help of X-ray analysis and DFT
calculations. This highlighted that the torsion on the aromatic
ring in the transition state significantly impacted the value of the
corresponding rotational barrier. The results described herein
provide general trends on the structure-rotational barrier relation-
ships which can be extrapolated to other atropisomeric series in
order to modulate their enantiomeric stability thanks to judicious
structural optimization. In particular, such data are very useful
during design and optimization of catalysts or drugs with a chiral
axis to end up with either a pair of fast exchanging rotamers or
either a pair of atropisomers.
3-(2-Chlorophenyl)-3,4,5,6-tetrahydro-2H-cyclohexa[d][1,3]thiazole-2-
thione 2b
Yield = 34%; Rf = 0.45 (petroleum ether/dichloromethane, 1/1);
mp 127.1 1C (racemate); 1H NMR (400 MHz, CDCl3) d 1.76–1.84
(4H, m, 2 CH2), 2.04–2.06 (2H, m, CH2), 2.51–2.53 (2H, m, CH2);
7.31–7.33 (1H, m, arom), 7.43–7.45 (2H, m, arom), 7.56–7.58
Experimental section
General information
Unless specified, all reagents, starting materials and solvents (1H, m, arom); 13C NMR (100 MHz, CDCl3) d 21.6, 22.6, 24.3,
were purchased from commercial sources and used as received. 26.9, 120.8, 128.1, 130.2, 130.7, 130.9, 132.5, 135.1, 136.9, 188.0;
Melting points are uncorrected. Analytical thin-layer chromato- HRMS (ESI/TOF) m/z [M + H]+ calcd for C13H13NS2Cl: 282.0172;
graphy (TLC) was performed using precoated silica gel plates found: 282.0174; chiral HPLC: Whelk-O1 (S,S), 25 1C, heptane/
and visualization was achieved by UV light (254 nm). Flash ethanol 60/40, 1 mL minꢀ1, UV and CD 254 nm, Rt1 = 6.71 min
chromatography was performed using silica gel and a gradient (+), Rt2 = 8.15 min (ꢀ), k1 = 1.24, k2 = 1.72, a = 1.39 and Rs = 3.76;
solvent system. 1H and 13C NMR spectra were measured on first eluted enantiomer (99.5% ee): [a]D25 ꢀ16 (c 0.30, CHCl3).
400 MHz spectrometer with CDCl3 as solvent. Chemical shifts
3-(2-Bromophenyl)-3,4,5,6-tetrahydro-2H-cyclohexa[d][1,3]thiazole-2-
thione 2c
(ppm) were recorded with respect to TMS in CDCl3. Multi-
plicities are given as s (singlet), d (doublet), t (triplet), q
(quartet), m (multiplet), dd (doublet of doublets) or ddd (doub- Yield: 40%; Rf = 0.36 (petroleum ether/dichloromethane, 1/1);
let of doublets of doublets). Coupling constants are reported as mp 132.0 1C (racemate); 1H NMR (400 MHz, CDCl3) d 1.79–1.86
a J value in Hz. HRMS data were recorded on a mass spectro- (4H, m, 2 CH2), 2.00–2.06 (2H, m, CH2), 2.52–2.54 (2H, m, CH2),
meter with electrospray ionization and TOF mass analyzer. 7.32–7.41 (2H, m, arom), 7.49–7.51 (1H, m, arom), 7.73–7.75
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
(1H, m, arom); 13C NMR (100 MHz, CDCl3) d 21.1, 22.1, 22.7, (100 MHz, CDCl3) d 16.1, 98.3, 106.7, 129.8, 130.0, 131.3, 139.3,
24.0, 120.2, 121.8, 128.4, 129.8, 130.6, 133.2, 136.2, 136.4, 186.9; 140.5, 140.8, 189.7; HRMS (ESI/TOF) m/z [M + H]+ calcd for
HRMS (ESI/TOF) m/z [M + H]+ calcd for C13H13NS2Br: 325.9667;
C10H9NS2I: 333.9216; found: 333.9220; chiral HPLC: Whelk-O1
found: 325.9666; chiral HPLC: Whelk-O1 (S,S), 25 1C, heptane/ (S,S), 25 1C, heptane/ethanol/dichloromethane 50/30/20,
ethanol 60/40, 1 mL minꢀ1, UV and polarimeter, Rt1 = 7.00 min 1 mL minꢀ1, UV and CD 254 nm, Rt1 = 4.69 min (ꢀ), Rt2
=
(ꢀ), Rt2 = 8.43 min (+), k1 = 1.33, k2 = 1.81, a = 1.36 and Rs = 3.60; 5.24 min (+), k1 = 0.59, k2 = 0.77, a = 1.32 and Rs = 2.68; first
first eluted (99% ee): [a]2D5 21 (c 0.79, CHCl3).
eluted (99% ee): [a]2D5 174 (c 0.47, CHCl3).
3-(2-Iodophenyl)-3,4,5,6-tetrahydro-2H-cyclohexa[d][1,3]thiazole-2- 3-(2-Iodophenyl)thiazole-2(3H)-thione 4d
Yield: 32%; Rf = 0.61 (dichloromethane); 1H NMR (400 MHz,
thione 2d
Yield: 48%; Rf = 0.43 (petroleum ether/dichloromethane, 1/1); CDCl3) d 6.71 (1H, d, J = 4.6, arom), 6.98 (1H, d, J = 4.6, arom),
mp 157.0 1C (racemate); 1H NMR (400 MHz, CDCl3) d 1.75–1.90 7.21 (1H, ddd, J = 1.5, 7.7, 7.9, arom), 7.40 (1H, dd, J = 1.4, 7.9,
(4H, m, 2 CH2), 2.00–2.11 (2H, m, CH2), 2.51–2.60 (2H, m, CH2), arom), 7.52 (1H, ddd, J = 1.4, 7.9, arom), 7.99 (1H, d, J = 1.1, 7.9,
7.20 (1H, dd, J = 1.6, 7.8, arom), 7.30 (1H, dd, J = 1.5, 7.8, arom), arom); 13C NMR (100 MHz, CDCl3) d 96.7, 111.7, 129.3, 129.8,
7.52 (1H, dd, J = 1.2, 7.6, arom), 7.98 (1H, dd, J = 1.5, 8.0, arom); 131.4, 131.5, 140.4, 141.5, 189.0; HRMS (ESI/TOF) m/z [M + H]+
13C NMR (100 MHz, CDCl3) d 21.7, 22.7, 23.3, 25.0, 98.0, 121.1, calcd for C9H7NS2I: 319.9059; found: 319.9057.
129.6, 129.7, 130.9, 136.5, 140.2, 140.5, 187.6; HRMS (ESI/TOF)
m/z [M + H]+ calcd for C13H13NS2I: 373.9529; found: 373.9532;
3-(2-Fluoro-6-methylphenyl)thiazole-2(3H)-thione 5a
chiral HPLC: Whelk-O1 (S,S), 25 1C, heptane/ethanol 60/40, Yield: 61%; Rf = 0.64 (dichloromethane); mp 117.0 1C (race-
1 mL minꢀ1, UV and polarimeter, Rt1 = 7.35 min (ꢀ), Rt2 = 8.94 mate); 1H NMR (400 MHz, CDCl3) d 2.23 (3H, s, CH3), 6.72 (1H,
min (+), k1 = 1.45, k2 = 1.98, a = 1.36 and Rs = 3.82; first eluted d, J = 4.7, arom), 6.95 (1H, d, J = 4.5, arom), 7.10 (1H, dd, J = 8.4,
(99.5% ee): [a]2D5 86 (c 0.73, CHCl3).
9.3, arom), 7.15 (1H, d, J = 7.7, arom), 7.38 (1H, ddd, J = 5.6, 7.7,
8.2, arom); 13C NMR (100 MHz, CDCl3) d 17.7 (d, J = 2.3), 111.9,
114.3 (d, J = 19.4), 125.7 (d, J = 13.3), 126.5 (d, J = 3.5), 131.1
3-(2-Chlorophenyl)-4-methylthiazole-2(3H)-thione 3b
Yield: 80%; Rf = 0.76 (dichloromethane/AcOEt, 9/1); mp (d, J = 8.3), 131.4, 138.6, 157.7 (d, J = 253.4), 189.2; HRMS (ESI/
1
150–152 1C (racemate); H NMR (400 MHz, CDCl3) d 1.94 (3H, TOF) m/z [M + H]+ calcd for C10H9NFS2: 226.0155; found:
s, CH3), 6.36 (1H, s, CH), 7.31 (1H, m, arom), 7.45–7.52 (2H, m, 226.0152; chiral HPLC: chiralpak IE, 25 1C, heptane/ethanol
arom), 7.59–7.64 (1H, m, arom); 13C NMR (100 MHz, CDCl3) 90/10, 1 mL minꢀ1, UV and CD 254 nm, Rt1 = 10.30 min (ꢀ),
d 15.6, 106.5, 128.5, 130.5, 131.0, 131.3, 132.8, 135.6, 139.6, Rt2 = 11.73 min (+), k1 = 2.43, k2 = 2.91, a = 1.20 and Rs = 3.77;
190.2; HRMS (ESI/TOF) m/z [M + H]+ calcd for C10H9NS2Cl: first eluted (99.5% ee): [a]D25 ꢀ148 (c 1.11, CHCl3).
241.9859; found: 241.9855; chiral HPLC: Whelk-O1 (S,S), 25 1C,
General synthesis of N-aryl-thiazoline-2-selenone 6
heptane/ethanol/dichloromethane 50/30/20,
1 ,
mL minꢀ1
UV and CD 254 nm, Rt1 = 4.43 min (+), Rt2 = 4.97 min (ꢀ), Iodomethane (7 mmol) was added under nitrogen atmosphere
k1 = 0.50, k2 = 0.68, a = 1.36 and Rs = 2.62; first eluted (99% ee): to a solution of N-aryl-thiazoline-2-thione (1.40 mmol) in acet-
[a]2D5 25 (c 0.51, CHCl3).
one (6.5 mL). The mixture was stirred 2–3 h at r.t. Then the
precipitate was filtered, washed with acetone and dried to give
the thiazolium salt. This salt (0.37 mmol) was used without any
3-(2-Bromophenyl)-4-methylthiazole-2(3H)-thione 3c
Yield: 46%; Rf = 0.76 (dichloromethane/AcOEt, 9/1); mp further purification and immediately solubilised in acetonitrile
163–165 1C (racemate); 1H NMR (400 MHz, CDCl3) d 1.94 (3 mL). This mixture was added dropwise to a solution of
(3H, d, J = 1.1, CH3), 6.36 (1H, q, J = 1.1, CH), 7.33 (1H, dd, selenium powder (0.73 mmol) and sodium borohydride
J = 1.5, 7.8, arom), 7.41 (1H, ddd, J = 1.6, 7.6, 7.9, arom), 7.53 (0.81 mmol) in anhydrous ethanol (13 mL). After stirring for
(1H, ddd, J = 1.3, 7.6, 7.7, arom), 7.79 (1H, dd, J = 1.2, 8.0, arom); 30 min, a solution of acetic acid in water (5 mol%) was added
13C NMR (100 MHz, CDCl3) d 15.8, 106.5, 122.9, 129.9, 130.5, dropwise, the precipitate was filtered off and washed with
131.5, 134.3, 137.3, 139.5, 190.1; HRMS (ESI/TOF) m/z [M + H]+ dichloromethane. The aqueous layer was separate from the
calcd for C10H9NS2Br: 287.9333; found: 287.9333; chiral HPLC: filtrate and extracted with dichloromethane (3 ꢁ 50 mL). The
Whelk-O1 (S,S), 25 1C, heptane/ethanol/dichloromethane combined organic layers were washed with water (3 ꢁ 50 mL),
50/30/20, 1 mL minꢀ1, UV and CD 254 nm, Rt1 = 4.53 min (+), dried on MgSO4 and evaporated under reduced pressure. The
Rt2 = 5.07 min (ꢀ), k1 = 0.54, k2 = 0.72, a = 1.34 and Rs = 2.66; desired product was purified by flash chromatography (petro-
first eluted (99% ee): [a]2D5 85 (c 0.44, CHCl3).
leum ether–dichloromethane, 100/0 - 0/100).
3-(2-Iodophenyl)-4-methylthiazole-2(3H)-thione 3d
3-(2-Fluorophenyl)-3,4,5,6-tetrahydro-2H-cyclopenta[d][1,3]thiazole-
2-selenone 6a
Yield: 49%; Rf = 0.76 (dichloromethane/AcOEt, 9/1); mp
1
193–195 1C (racemate); H NMR (400 MHz, CDCl3) d 1.93 (3H, Yield: 32%; brown solid; mp 163.0–165.0 1C (racemate);
d, J = 1.1, CH3), 6.37 (1H, q, J = 1.1, CH), 7.23 (1H, ddd, J = 1.5, 1H NMR (400 MHz, CDCl3) d 2.49–2.62 (4H, m, 2 CH2),
7.6, 7.9, arom), 7.30 (1H, dd, J = 1.4, 7.8, arom), 7.56 (1H, ddd, 2.75–2.81 (2H, m, CH2), 7.27–7.34 (2H, m, arom), 7.44–7.54
J = 1.3, 7.6, 7.8, arom), 8.01 (1H, dd, J = 1.2, 8.0, arom); 13C NMR (2H, m, arom); 13C NMR (100 MHz, CDCl3) d 26.1, 27.1, 28.1,
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem.

View Article Online
Paper
NJC
117.2 (d, J = 19.0), 124.9 (d, J = 3.7), 126.3 (d, J = 12.4), 128.2, atmosphere. The mixture was stirred 12 h at r.t. Water was
129.7, 131.6 (d, J = 7.5), 148.8, 156.6 (d, J = 253.0), 187.8; 19F added (100 mL) and the mixture was extracted with ethyl
NMR (376.5 MHz, CDCl3) d ꢀ120.0; HRMS (ESI/TOF) m/z acetate (3 ꢁ 100 mL), the organic layer was dried on MgSO4
[M + H]+ calcd for C12H11NFSSe: 299.9756; found: 299.9756.
and evaporated under reduced pressure. The desired product
was purified by flash chromatography (dichloromethane–ethyl
acetate, 100/0 - 60/40).
3-(2-Chlorophenyl)-3,4,5,6-tetrahydro-2H-cyclopenta[d][1,3]thiazole-
2-selenone 6b
Yield: 9%; brown oil; 1H NMR (400 MHz, CDCl3) d 2.45–2.58
(4H, m, 2 CH2), 2.77–2.82 (2H, m, CH2), 7.43–7.62 (4H, m,
3-(2-Fluorophenyl)-3,4,5,6-tetrahydro-2H-cyclopenta[d][1,3]thiazol-
2-one 7a
arom); 13C NMR (100 MHz, CDCl3) d 26.4, 27.3, 28.3, 128.3, Yield: 34%; colorless oil; 1H NMR (400 MHz, CDCl3) d 2.22–2.32
128.8, 129.8, 131.0, 131.3, 131.7, 136.4, 148.8, 186.7; HRMS (2H, m, CH2), 2.47–2.57 (2H, m, CH2), 2.75–2.84 (2H, m, CH2),
(ESI/TOF) m/z [M + H]+ calcd for C12H11NSClSe: 315.9458; 7.19–7.27 (2H, m, arom), 7.32–7.42 (2H, m, arom); 13C NMR
found: 315.9462; chiral HPLC: Whelk-O1 (S,S), 10 1C, (100 MHz, CDCl3) d 23.4, 27.8 (d, J = 2.0 Hz), 29.3, 112.3,
heptane/ethanol 60/40, 1 mL minꢀ1, UV and polarimeter, 117.0 (d, J = 19.8), 124.0 (d, J = 13.1), 124.9 (d, J = 4.3), 129.4,
Rt1 = 6.96 min (+), Rt2 = 8.07 min (ꢀ), k1 = 1.32, k2 = 1.69, 130.5 (d, J = 7.7), 136.6, 157.6 (d, J = 251.0), 175.0; 19F NMR
a = 1.28 and Rs = 3.44; first eluted (99.5% ee): [a]2D5 ꢀ91 (c 0.15, (376,5 MHz, CDCl3) d ꢀ120.1; HRMS (ESI/TOF) m/z [M + H]+
CHCl3).
calcd for C12H11NOSF: 236.0540; found: 236.0541; chiral HPLC:
Whelk-O1 (S,S), 10 1C, heptane/ethanol/dichloromethane
70/20/10, 1 mL minꢀ1, UV and polarimeter, Rt = 7.81 min.
3-(2-Bromophenyl)-3,4,5,6-tetrahydro-2H-cyclopenta[d][1,3]
thiazole-2-selenone 6c
3-(2-Chlorophenyl)-3,4,5,6-tetrahydro-2H-cyclopenta[d][1,3]thiazol-2-
one 7b
Yield: 32%; yellow solid; mp 174.5–176.2 1C (racemate);
1H NMR (400 MHz, CDCl3) d 2.40–2.60 (4H, m, 2 CH2),
2.75–2.82 (2H, m, CH2), 7.36–7.43 (2H, m, arom), 7.46–7.54 Yield: 43%; colorless oil; 1H NMR (400 MHz, CDCl3) d 2.15–2.24
(1H, m, arom), 7.74–7.77 (1H, m, arom); 13C NMR (100 MHz, (2H, m, CH2), 2.30–2.46 (2H, m, CH2), 2.65–2.79 (2H, m, CH2),
CDCl3) d 26.2, 27.3, 28.2, 121.4, 128.2, 128.9, 129.8, 131.3, 134.0, 7.24–7.37 (3H, m, arom), 7.43–7.49 (1H, m, arom); 13C NMR
138.1, 148.4, 187.1; HRMS (ESI/TOF) m/z [M + H]+ calcd for (100 MHz, CDCl3) d 23.3, 27.7, 29.2, 112.0, 127.8, 129.8, 130.3,
C
12H11NSBrSe: 359.8952; found: 359.8948; chiral HPLC: Whelk- 130.6, 132.6, 133.9, 136.3, 174.8; HRMS (ESI/TOF) m/z [M + H]+
O1 (S,S), 25 1C, heptane/ethanol 60/40, 1 mL minꢀ1, UV and CD calcd for C12H11NOSCl: 252.0244; found: 252.0247; chiral
254 nm, Rt1 = 7.93 min (ꢀ), Rt2 = 9.14 min (+), k1 = 1.69, HPLC: Whelk-O1 (S,S), 25 1C, heptane/ethanol 60/40, 1 mL
k2 = 2.10, a = 1.24 and Rs = 3.10; first eluted (99.5% ee): minꢀ1, UV and polarimeter, Rt1 = 5.93 min, Rt2 = 8.69 min,
[a]2D5 ꢀ22 (c 0.18, CHCl3).
k1 = 0.98, k2 = 1.90, a = 1.94 and Rs = 7.65; first eluted (99.5% ee):
[a]2D5 7 (c 0.28, CHCl3).
3-(2-Iodophenyl)-3,4,5,6-tetrahydro-2H-cyclopenta[d][1,3]thiazole-2-
selenone 6d
3-(2-Bromophenyl)-3,4,5,6-tetrahydro-2H-cyclopenta[d][1,3]thiazol-2-
one 7c
Yield: 11%; yellow solid; mp 184.5–186.8 1C (racemate);
1
1H NMR (400 MHz, CDCl3) d 2.39–2.58 (4H, m, 2 CH2), Yield: 36%; brown oil; H NMR (400 MHz, CDCl3) d 2.19–2.32
2.73–2.85 (2H, m, CH2), 7.23 (1H, dd, J = 1.6, 7.8, arom), 7.36 (2H, m, CH2), 2.34–2.54 (2H, m, CH2), 2.77–2.83 (2H, m, CH2),
(1H, dd, J = 1.6, 8.0, arom), 7.53 (1H, dd, J = 1.3, 7.7, arom), 8.0 7.29–7.34 (2H, m, arom), 7.40–7.44 (1H, m, arom), 7.71 (1H, dd,
(1H, dd, J = 1.6, 8.0, arom); 13C NMR (100 MHz, CDCl3) d 26.5, J = 1.2, 8.0, arom); 13C NMR (100 MHz, CDCl3) d 23.3, 27.7, 29.2,
27.8, 28.4, 96.8, 128.7, 129.2, 130.0, 131.4, 140.5, 141.8, 148.3, 112.0, 122.8, 128.6, 130.0, 130.7, 133.8, 135.7, 136.2, 174.8;
187.0; HRMS (ESI/TOF) m/z [M + H]+ calcd for C12H11NSISe: HRMS (ESI/TOF) m/z [M + Na]+ calcd for C12H10NOSBrNa:
407.8816; found: 407.8814; chiral HPLC: Whelk-O1 (S,S), 25 1C, 319.9538; found: 319.9539; chiral HPLC: Whelk-O1 (S,S),
heptane/ethanol 60/40, 1 mL minꢀ1, UV and CD 254 nm, 25 1C, heptane/ethanol 60/40, 1 mL minꢀ1, UV and polarimeter,
Rt1 = 8.25 min (ꢀ), Rt2 = 9.53 min (+), k1 = 1.80, k2 = 2.23, a = Rt1 = 6.52 min (+), Rt2 = 9.80 min (ꢀ), k1 = 1.17, k2 = 2.27,
1.24 and Rs = 3.58; first eluted (99.5% ee): [a]2D5 26 (c 0.07, a = 1.94 and Rs = 8.07; first eluted (99.5% ee): [a]D25 16 (c 0.96,
CHCl3).
CHCl3).
General synthesis of N-aryl-thiazoline-2-one 7
3-(2-Iodophenyl)-3,4,5,6-tetrahydro-2H-cyclopenta[d][1,3]thiazol-2-
one 7d
Iodomethane (7.50 mmol) was added under nitrogen atmo-
sphere to a solution of an N-aryl-thiazoline-2-thione (1.50 mmol) Yield: 67%; yellow solid; mp 148.5–150.2 1C (racemate);
in acetone (5 mL). The mixture was stirred 2–3 h at r.t. Then the 1H NMR (400 MHz, CDCl3) d 2.20–2.30 (2H, m, CH2), 2.30–
precipitate was filtered, washed with acetone and dried to give 2.41 (1H, m, CH2), 2.41–2.52 (1H, m, CH2), 2.72–2.87 (2H, m,
the thiazolium salt. This salt (0.61 mmol) was used without any CH2), 7.14 (1H, ddd, J = 1.6, 7.6, 8.0, arom), 7.29 (1H, dd, J = 1.5,
further purification and immediately solubilised in methanol 7.8, arom), 7.45 (1H, ddd, J = 1.3, 7.5, 7.9, arom), 7.93 (1H, dd,
(6.60 mL). A solution of sodium methoxide (0.61 mmol) in J = 1.2, 8.0, arom); 13C NMR (100 MHz, CDCl3) d 23.3, 27.8, 29.6,
methanol (24 mL) was then added at r.t. under nitrogen 98.1, 112.1, 129.3, 129.4, 130.7, 135.8, 139.2, 139.9, 174.6;
New J. Chem.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
HRMS (ESI/TOF) m/z [M + H]+ calcd for C12H11NOSI: 343.9601;
found: 343.9600; chiral HPLC: Whelk-O1 (S,S), 25 1C, heptane/
ethanol 60/40, 1 mL minꢀ1, UV and polarimeter, Rt1 = 6.87 min
(+), Rt2 = 9.98 min (ꢀ), k1 = 1.29, k2 = 2.33, a = 1.81 and Rs = 7.00;
first eluted (99.5% ee): [a]2D5 35 (c 0.93, CHCl3).
8 V. Belot, D. Farran, M. Jean, M. Albalat, N. Vanthuyne and
C. Roussel, J. Org. Chem., 2017, 82, 10188–10200.
9 (a) G. Bott, L. D. Field and S. Sternhell, J. Am. Chem. Soc.,
1980, 102, 5618–5626; (b) R. Ruzziconi, S. Spizzichino,
L. Lunazzi, A. Mazzanti and M. Schlosser, Chem. – Eur. J.,
¨
2009, 15, 2645–2652; (c) O. D. Ordu and I. Dogan, Tetrahe-
dron: Asymmetry, 2004, 15, 925–933.
Conflicts of interest
10 N. Vanthuyne, F. Andreoli, S. Fernandez, M. Roman and
C. Roussel, Lett. Org. Chem., 2005, 2, 433–443.
11 O. Trapp and V. Schurig, Chirality, 2002, 14, 465–470.
12 DFT calculated energies and the computational details are
given in the ESI†.
There are no conflicts to declare.
Acknowledgements
13 For a study about the influence of the ring size on the
rotational barrier of an atropisomers series see: J. E. Diaz,
A. Mazzanti, L. R. Orelli and M. Mancinelli, ACS Omega,
2019, 4, 4712–4720.
We thank Dr M. Giorgi (Spectropole) for X-ray structures, and
´
´
´
´
the ‘‘Centre Regional de Competences en Modelisation Mole-
culaire de Marseille’’ for computing facilities.
14 Plots of experimental versus calculated values are available
in the ESI†.
References
15 Same trend observed with three different functional/basis
sets (B3LYP/6-311G(3d,3p), M06-2X/6-311G(3d,3p) and
PBE0/def2-TZVPP): see ESI†.
16 Attempts of enantiomers separation performed on the
following immobilized CSPs: Chiralpak IA, Chiralpak IB,
Chiralpak IC, Chiralpak ID, Chiralpak IE and Chiralpak IF.
17 W. H. Pirkle, M. E. Koscho and Z. Wu, J. Chromatogr. A,
1996, 726, 91–97.
1 IUPAC, Compendium of Chemical Terminology, version 2.3.3,
2014, 126–270.
2 (a) M. Mancinelli, G. Bencivenni, D. Pecorari and
A. Mazzanti, Eur. J. Org. Chem., 2020, 4070–4086;
(b) E. Kumarasamy, R. Raghunathan, M. P. Sibi and
J. Sivaguru, Chem. Rev., 2015, 115, 11239–11300.
3 (a) M. Berthod, G. Mignani, G. Woodward and M. Lemaire,
Chem. Rev., 2005, 105, 1801–1836; (b) Y. Chen, S. Yekta and
A. K. Yudin, Chem. Rev., 2003, 103, 3155–3211.
18 See ESI†.
19 (a) N. Bellec, D. Lorcy and A. Robert, Synthesis, 1998,
1442–1446; (b) N. Bellec, D. Lorcy, A. Robert, R. Carlier,
A. Tallec, C. Rimbaud, L. Ouahab, R. Clerac and P. Delhaes,
Adv. Mater., 1997, 9, 1052–1056.
20 C. Roussel, M. Adjimi, A. Chemlal and A. Djafri, J. Org.
Chem., 1988, 53, 5076–5080.
21 A. Bondi, J. Phys. Chem., 1964, 68, 441–451.
22 Such structural modifications on atropisomers have already
been reported in literature: (a) E. M. Yilmaz and I. Dogan,
4 (a) M. M. Cardenas, A. D. Nguyen, Z. E. Brown, B. S. Heydari,
B. S. Heydari, S. D. Vaidya and J. L. Gustafson, ARKIVOC,
2021, 20–47; (b) S. T. Toenjes and J. L. Gustafson, Future
Med. Chem., 2018, 10, 409–422; (c) P. W. Glunz, Bioorg. Med.
Chem. Lett., 2018, 28, 53–60; (d) J. Chandrasekhar, R. Dick,
J. Van Veldhuizen, D. Koditek, E.-I. Lepist, M. E. McGrath,
L. Patel, G. Phillips, K. Sedillo, J. R. Somoza, J. Therrien,
N. A. Till, J. Treiberg, A. G. Villasenor, Y. Zherebina and
S. Perrault, J. Med. Chem., 2018, 61, 6858–6868;
(e) D. E. Smith, I. Marquez, M. E. Lokensgard,
A. L. Rheingold, D. A. Hecht and J. L. Gustafson, Angew.
Chem., Int. Ed., 2015, 54, 11754–11759; ( f ) J. E. Smyth,
N. M. Butler and P. A. Keller, Nat. Prod. Rep., 2015, 32,
1562–1583.
5 (a) S. R. LaPlante, P. J. Edwards, L. D. Fader, A. Jakalian and
O. Hucke, ChemMedChem, 2011, 6, 505–513; (b) S. R.
LaPlante, L. D. Fader, K. R. Fandrick, D. R. Fandrick,
O. Hucke, R. Kemper, S. P. F. Miller and P. J. Edwards,
J. Med. Chem., 2011, 54, 7005–7022; (c) J. Clayden,
W. J. Moran, P. J. Edwards and S. R. LaPlante, Angew. Chem.,
Int. Ed., 2009, 48, 6398–6401.
¨
Tetrahedron: Asymmetry, 2008, 19, 2184–2191; (b) O. Demir-
Ordu, E. M. Yilmaz and I. Dogan, Tetrahedron: Asymmetry,
2005, 16, 3752–3761; (c) F. Oguz and I. Dogan, Tetrahedron:
Asymmetry, 2003, 14, 1857–1864.
23 None enantiomers separation observed on the Whelk-O1
(S,S) column at 10 1C (heptane/ethanol/dichloromethane –
70/20/10).
24 van der Waals radius for oxygen atom = 142 pm; see
ref 21.
25 Same trend observed with three different functional/basis
sets (B3LYP/6-311G(3d,3p), M06-2X/6-311G(3d,3p) and
PBE0/def2-TZVPP): see ESI†.
26 A. Iida, M. Matsuoka, H. Hasegawa, N. Vanthuyne,
D. Farran, C. Roussel and O. Kitagawa, J. Org. Chem.,
2019, 84, 3169–3175.
6 M. Oki, Top. Stereochem., 1983, 14, 1–81.
7 I. Alkorta, J. Elguero, C. Roussel, N. Vanthuyne and P. Piras,
Adv. Heterocycl. Chem., 2012, 105, 1–188.
This journal is © The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem.