Stereoselectivity in dehydrative cyclic trimerization of substituted 4-alkylaminobenzoic acids

Article abstract of DOI:10.1039/d0nj05368f

The cyclic trimerization of substituted 4-alkylaminobenzoic acids was investigated. From NMR analyses of DiMeO_C3A with two methoxy groups, which was obtained by using SiCl4 as a dehydrative condensation reagent and purified by preparative GPC, a syn/anti

Full text of DOI:10.1039/d0nj05368f

NJC
View Article Online
View Journal | View Issue
PAPER
Stereoselectivity in dehydrative cyclic
trimerization of substituted 4-alkylaminobenzoic
acids†
Cite this: New J. Chem., 2021,
45, 1187
Koji Takagi, *a Hinako Yamaguchi,a Daiki Miyamoto,a Yuka Deguchi,a
b
Takehiro Hiraob and Takeharu Haino
The cyclic trimerization of substituted 4-alkylaminobenzoic acids was investigated. From NMR analyses
of DiMeO_C3A with two methoxy groups, which was obtained by using SiCl4 as a dehydrative
condensation reagent and purified by preparative GPC, a syn/anti ratio of 60/40 was obtained. On the
other hand, 3Br_C3A with one bromine group at the ortho-position relative to the amide nitrogen was
synthesized by using PPh3/Cl3CCCl3 as a dehydrative condensation reagent and isolated by SiO2 column
chromatography. 3Br_C3A showed an inverse stereoselectivity, namely, a syn/anti ratio of 25/75 was
calculated based on the comprehensive NMR analyses. The population of stereoisomers had no
relationship with the dehydrative condensation reagent and reaction temperature. The solvent character
also had a negligible influence on the syn/anti ratio in solution reflecting the rigid structure of 3Br_C3A.
Received 2nd November 2020,
Accepted 3rd December 2020
DOI: 10.1039/d0nj05368f
rsc.li/njc
calix[4]arene scaffolds have been applied to organize perylene
bisimide chromophores in zigzag-type arrangements providing
Introduction
Since the accidental preparation of crown ethers and their defined distances and angles between chromophores.12,13 The
complexation with alkali and alkaline earth metal ions devel- zigzag-type arrangement of dyes gave rise to the efficient sequen-
oped by Pedersen in 1967,1 many synthetic macrocyclic com- tial energy transfer process along the oligomeric chain.
pounds have been investigated due to the specific and selective
Calix[3]amide classified by Azumaya and co-workers is a
host–guest complexation. In addition to the fundamental cyclic trimer having a meta-linked N-alkylated benzanilide
chemistry related to the host–guest interaction, this molecular skeleton14,15 and is efficiently synthesized by the condensation
recognition event stimulated the research activity associated reaction of 3-alkylaminobenzoic acid derivatives due to the
with supramolecular chemistry, organic catalysis, polymer preferential cis conformation of N-alkylated benzanilides.16
formation, and so on. From this point of view, cyclic oligomers Calix[3]amides having functional groups on the benzene ring
including calixarenes,2 resorcinarenes,3 cucurbiturils,4,5 cyclo- were also synthesized.17–19 We have achieved the chiral three
dextrins6,7 and pillararenes8–10 with unique cavities have dimensional alignment of p-conjugated chromophores with triple-
been drawing much attention in the past few decades. stranded helicity, in which three bithiophene chromophores are
The design and synthesis of new cyclic oligomers is of high sandwiched by two calix[3]amides.20 However, the helicity of
importance and allows chemists to explore novel science.11 bithiophenes was dynamic to only show the strong Cotton effect
Furthermore, cyclic oligomers having a well-defined conformation in selected solvents or at low temperature. Subsequently, we
can provide the molecular system for studying the through-space have reported the synthesis and chiroptical properties of new
interactions between segments to which the cyclic oligomer is cyclic tri(benzamide) carrying bis(phenylethynyl)benzene chro-
covalently bound. Cyclic oligomers can also serve as a building mophores arranged in a triple-stranded helical fashion.21 It is
block to construct supramolecular architectures leading to the to be noted that the diastereoselective synthesis of cyclic
development of materials with smart functions. For example, tri(benzamide) having six bromo groups (DiBr_C3A) proceeds
exclusively (Fig. 1).
a Graduate School of Engineering, Nagoya Institute of Technology, Gokiso-cho,
Showa-ku, Nagoya, Aichi 466-8555, Japan. E-mail: takagi.koji@nitech.ac.jp
b Graduate School of Advanced Science and Engineering, Hiroshima University,
1-3-1 Kagamiyama, Higashi-Hiroshima, Hiroshima 739-8526, Japan
† Electronic supplementary information (ESI) available: Synthetic schemes,
GPC chromatograms, ESI-MS charts, NMR spectra, and computational results.
See DOI: 10.1039/d0nj05368f
Because the introduction of functional groups is highly
demanded for the post-modification of cyclic oligomers (vide
supra), our previous finding gave invaluable information about
the stereoselective cyclization of 4-alkylaminobenzoic acids.
On the other hand, the role of the bromine group is not fully
understood and the incorporation of other substituents has not
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 1187--1193 | 1187

View Article Online
Paper
NJC
The resulting DOSY data were analyzed using the MestReNova
program to obtain the diffusion coefficient values discussed
in the main text. The signal decay of selected aromatic
protons was fit to a mono-exponential fitting function
(eqn (1)).
I = I0eꢀxD
(1)
I and I0 denote the NMR signal intensities in the presence and
absence of gradient pulses, respectively. D is the diffusion
coefficient value. The x indicates [ꢀg2g2d2(D–d/3)], where g, g,
d, and D denote the gyromagnetic ratio, gradient strength, its
duration gradient, and separation between the edges of the
gradient pulses, respectively.22
Fig. 1 Dehydrative cyclic trimerization of 4-alkylaminobenzoic acids.
been investigated in detail. In this contribution, we describe
the stereoselectivity in the dehydrative cyclic trimerization of
two monomers to address this issue (Fig. 1). The comprehensive
NMR analyses and preliminary theoretical calculations are
reported.
Materials
All materials were obtained from commercial suppliers and
used as received. Anhydrous tetrahydrofuran (THF) and
dichloromethane (DCM) were purchased from Kanto Chemical.
Other solvents were dried and distilled following the standard
methods, and stored under a nitrogen atmosphere.
Experimental
Instruments
Synthesis
Microwave-assisted reactions were performed in a Biotage
microwave reactor (Initiator) using 0.5–2.0 mL microwave reaction
vials. Melting points (Mps) were determined on a Yanaco micro
melting point apparatus MP-500D. Elemental analyses (EA) were
performed on an Elementar vario EL cube in the CHN mode.
Gel permeation chromatography (GPC) analyses were carried out
with a Shodex 104 system using tandem LF-404 columns (THF as
the eluent, flow rate = 1.0 mL minꢀ1, 40 1C) equipped with
refractive index (RI) and ultraviolet (UV) detectors. Purifications
with preparative GPC were carried out on a Japan analytical
industry LC-9210 system using tandem JAIGEL 1H, 2H, and
2,5-Dimethoxy-4-nitrobenzoic acid methyl ester (1). A mixture
of 2,5-dihydroxy-4-nitrobenzoic acid methyl ester (2.77 g,
13.1 mmol),23 K2CO3 (3.98 g, 28.8 mmol), and iodomethane
(1.80 mL, 28.8 mmol) in acetone (80 mL) was heated to reflux
for 48 h under a nitrogen atmosphere. After cooling, acetone
was evaporated and then ethyl acetate and water were added.
An aqueous phase was extracted with ethyl acetate, and the
combined organic phase was rinsed with brine and dried over
MgSO4. After evaporating solvents, the obtained crude product
was used for the next reaction without purification (2.95 g,
99%). 1H-NMR (CDCl3, ppm) 3.91 (s, 3H), 3.94 (s, 3H), 3.96
(s, 3H), 7.45 (s, 1H), 7.51 (s, 1H).
2.5H columns (CHCl3 as an eluent, flow rate = 3.8 mL minꢀ1
)
equipped with an UV detector monitored at 254 nm. High
resolution electrospray ionization mass spectra (HR ESI-MS) were
obtained on a Waters Synapt G2 HDMS (Nagoya Institute of
Technology, NIT) and Thermo Fisher Scientific LTQ Orbitrap XL
(Hiroshima Univ). One dimensional NMR spectroscopy and
nuclear Overhauser effect spectroscopy (NOESY) were performed
on a Bruker Avance III HD NanoBay 400 FT-NMR spectrometer
(NIT). Heteronuclear single quantum correlation (HSQC), double
quantum filtered COSY (DQF-COSY), diffusion ordered spectro-
scopy (DOSY), and valuable temperature NMR (VT-NMR) were
performed on a JEOL JNM-ECA500 spectrometer (Hiroshima
Univ). Theoretical calculations were performed using Gaussian
09 (Revision E.01) package of programs at Research Center
for Computational Science, Okazaki, Japan. The ground state
structures were optimized with density functional theory (DFT)
calculations at the B3LYP/6-31G(d) level of theory.
4-Amino-2,5-dimethoxybenzoic acid methyl ester (2). A CH3OH
solution (35 mL) of SnCl2 (27.5 g, 140 mmol) was added
dropwise to 1 (6.60 g, 29.3 mmol) in CH3OH/THF mixed
solution (70 mL, 1/1 in volume), and the mixture was heated
to reflux overnight. After neutralizing with saturated Na2CO3
solution, an aqueous phase was extracted with DCM. The
combined organic phase was rinsed with saturated NaHCO3
solution and dried over MgSO4. The crude product was purified
by SiO2 column chromatography (DCM : acetone = 6 : 1, Rf =
0.70) to give a brown solid (4.10 g, 67%). 1H-NMR (CDCl3, ppm)
3.83 (s, 6H), 3.85 (s, 3H), 4.25 (br, 2H), 6.30 (s, 1H), 7.34 (s, 1H).
2,5-Dimethoxy-4-hexylaminobenzoic acid methyl ester (3). After
dissolving 2 (4.10 g, 19.5 mmol), hexanal (3.66 mL, 30.0 mmol),
and AcOH (0.74 mL, 12.9 mmol) in THF (90 mL), NaBH(OAc)3
(7.50 g, 35.3 mmol) was added and the mixture was stirred at
room temperature overnight. After adding the saturated
Methods
Chloroform-d1 solution of Br_C3A (0.03 mmol mLꢀ1) was NaHCO3 solution, an aqueous phase was extracted with DCM.
placed in an NMR sample tube (3 mmf). The pulse-field The combined organic phase was rinsed with brine and dried
gradient diffusion NMR spectra were recorded using a bipolar over MgSO4. The crude product was purified by SiO2 column
pulse pair stimulated echo pulse sequence on a JEOL JNM-ECA500 chromatography (ethyl acetate : hexane = 2 : 3, Rf = 0.35) to give
1
spectrometer with a three mm inverse H3X/FG probe at 24 1C. brown oil (3.22 g, 56%). H-NMR (CDCl3, ppm) 0.89 (br, 3H),
1188 | New J. Chem., 2021, 45, 1187--1193
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
1.21–1.67 (8H), 3.16 (t, J = 6.82 Hz, 2H), 3.84 (s, 6H), 3.89 (s, 3H), After evaporating pyridine, 1 M HCl was added and an aqueous
6.12 (s, 1H), 7.29 (s, 1H).
phase was extracted with DCM. The combined organic phase
was rinsed with brine and dried over MgSO4. The crude product
was purified by SiO2 column chromatography (DCM, Rf = 0.44)
to give a pale yellow gummy solid (57 mg, 30%). 1H-NMR
(CDCl3, ppm) 0.87 (9H), 1.25 (30H), 1.65 (6H), 3.18 (3H), 4.23
(3H), 6.68–7.64 (9H). 13C-NMR (CDCl3, ppm) 14.1, 22.6, 26.9,
27.1, 27.3, 27.5, 27.6, 29.1, 29.2, 29.7, 31.7, 47.5, 47.6, 47.8,
123.2, 123.4, 123.9, 124.0, 124.3, 124.4, 126.3, 126.8, 130.7,
131.1, 131.7, 131.9, 132.0, 132.3, 132.8, 137.8, 138.1, 138.2,
138.4, 141.1, 141.2, 168.8 (some signals are overlapped).
HR-ESI-MS calcd for C45H60N3O3Br3 + H: 928.2258. Found:
928.2266.
2,5-Dimethoxy-4-hexylaminobenzoic acid (4). A mixture of 3
(3.22 g, 10.9 mmol) and 2 M NaOH solution (45 mL) in CH3OH
(100 mL) was heated at 40 1C for 24 h. After concentrating, 2 M
HCl solution was added to acidify the solution. The precipitated
solid was collected by filtration and washed with water to give
a pale yellow solid (1.68 g, 56%). M.p. 88–90 1C. 1H-NMR
(CDCl3, ppm) 1.35 (6H), 1.43 (3H), 1.69 (2H), 3.18 (t, J = 6.82 Hz,
2H), 3.87 (s, 3H), 4.03 (s, 3H), 6.13 (s, 1H), 7.43 (s, 1H).
13C-NMR (CDCl3, ppm) 14.0, 22.6, 26.8, 29.1, 31.5, 43.1, 55.9,
56.9, 92.1, 102.9, 111.9, 141.0, 144.4, 155.1, 166.3. Anal. calcd
for C15H23NO4: C, 64.03; H, 8.24; N, 4.98%. Found: C, 63.90; H,
8.48; N, 4.71%.
3Br_C3Ared. To LiAlH4 (46 mg, 1.2 mmol) and AlCl3 (0.16 g,
1.2 mmol) in ether (2.4 mL) was added dropwise a THF solution
(1.2 mL) of 3Br_C3A (93 mg, 0.10 mmol) at 0 1C, and the mixture
was stirred for 2 h at room temperature. After quenching with
cold 2 M aq. NaOH, an aqueous phase was extracted with DCM.
The combined organic phase was washed with brine and dried
over MgSO4. The crude product was purified by preparative
GPC (CHCl3 as an eluent) to give pale yellow oil (45 mg, 51%).
1H-NMR (CDCl3, ppm) 0.86 (9H), 1.26 (30H), 1.65 (6H), 2.87–4.39
(12H), 6.42–7.57 (9H). 13C-NMR (CDCl3, ppm) 14.1, 22.6, 26.7,
26.8, 27.0, 27.2, 29.3, 29.5, 29.7, 31.8, 54.6, 55.1, 55.5, 56.2, 56.5,
56.9, 121.6, 123.2, 123.6, 124.8, 125.2, 126.0, 126.2, 126.8, 127.1,
128.3, 133.1, 133.3, 133.6, 133.7, 134.6, 135.3, 135.7, 145.2, 145.6.
(some signals are overlapped). HR-ESI-MS calcd for C45H66N3Br3
+ H: 886.2885. Found: 886.2863.
DiMeO_C3A. SiCl4 (230 mg, 1.34 mmol) was added to 4
(250 mg, 0.89 mmol) in pyridine (5 mL) at 0 1C, and the mixture
was heated at 120 1C under microwave irradiation for 24 h.
After evaporating pyridine, 2 M HCl was added and an aqueous
phase was extracted with DCM. The combined organic phase
was rinsed with brine and dried over MgSO4. The crude product
was purified by SiO2 column chromatography (ethyl acetate,
Rf = 0.85) to give viscous oil (189 mg, 27%). After purification
with preparative GPC, the target compound was isolated.
1H-NMR (CDCl3, ppm) 0.89 (9H), 1.26 (18H), 1.48 (6H), 3.28–
4.10 (24H), 6.17–6.75 (6H). HR-ESI-MS calcd for C45H63N3O9 +
Na: 812.4464. Found: 812.4438.
4-Octylaminobenzoic acid (5). NaCNBH3 (1.1 g, 15 mmol) was
added to an EtOH (55 mL) solution of p-aminobenzoic acid
(2.0 g, 15 mmol) and octanal (2.3 mL, 15 mmol) at 0 1C. The
mixture was stirred overnight while returning to room temperature.
1 M HCl solution was added to acidify the solution. The precipi-
tated solid was collected by filtration and washed with water.
The crude product was recrystallized from water/EtOH (1/4 in
volume) to give a colorless solid (3.2 g, 88%). 1H-NMR (CDCl3,
ppm) 0.85 (t, J = 6.60 Hz, 3H), 1.26 (br, 10H), 1.53 (m, 2 H),
3.03 (m, 2H), 6.41 (t, J = 5.26 Hz, 1H), 6.54 (d, J = 8.56 Hz, 2H), 7.65
(d, J = 8.56 Hz, 2H), 12.0 (br, 1H).
Results and discussion
In the first part of this paper, the stereoselectivity in the
dehydrative cyclic trimerization of 2,5-dimethoxy-4-hexylamino-
benzoic acid (4) was investigated (Scheme 1). The monomer was
synthesized in four steps starting from 2,5-dihydroxy-4-
nitrobenzoic acid methyl ester24 prepared by Kilbinger and
co-workers (Scheme S1, ESI†). Following our previous paper,23
the cyclization of 4 was carried out using SiCl4 as a dehydrative
condensation reagent in pyridine at 120 1C under microwave
heating for 24 h. After purifying by SiO2 column chromatography,
the product was analyzed by GPC to find out that higher
molecular weight oligomers (peak B) are contaminated to the
target compound (peak A) (Fig. S1, ESI†). A cyclic trimer
DiMeO_C3A was finally isolated by the preparative GPC and
the purity was unambiguously confirmed by the HR ESI-MS
measurement (Fig. S2, ESI†). The detected mass number
3-Bromo-4-octylaminobenzoic acid (6). A mixture of 5 (7.4 g,
30 mmol) and N-bromosuccinimide (5.4 g, 30 mmol) in DCM
(180 mL) was stirred at room temperature for 20 h. The pre-
cipitated solid was filtered off, and the solution was rinsed with
1 M HCl solution and the solvent was evaporated to dryness to
give a colorless solid (4.4 g, 46%). M.p. 120–122 1C. 1H-NMR
(CDCl3, ppm) 0.89 (t, J = 6.72 Hz, 3H), 1.29 (br, 10H), 1.69
(m, 2H), 3.23 (br, 2H), 4.86 (br, 1H), 6.60 (d, J = 8.56 Hz, 1H), 7.92
(dd, J = 8.56, 1.96 Hz, 1H), 8.17 (d, J = 1.96 Hz, 1H). 13C-NMR
(CDCl3, ppm) 14.1, 22.7, 27.0, 29.0, 29.2, 29.3, 31.8, 43.5,
108.4, 109.5, 117.5, 131.5, 134.7, 149.1, 171.3. Anal. calcd for
C15H22BrNO2: C, 54.89; H, 6.76; N, 4.27%. Found: C, 54.86;
H, 6.76; N, 4.19%.
3Br_C3A. A mixture containing 6 (0.20 g, 0.61 mmol), PPh3
(0.19 g, 0.73 mmol), Cl3CCCl3 (0.17 g, 0.73 mmol) in pyridine
(4.6 mL) was heated at 120 1C under microwave irradiation for 24 h.
Scheme 1 Cyclic trimerization of 2,5-dimethoxy-4-hexylaminobenzoic
acid (4).
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 1187--1193 | 1189

View Article Online
Paper
NJC
Fig. 3 Population of syn/anti stereoisomers in DiMeO_C3A.
in the bottom right chart of Fig. 2. More importantly, similar
cross-peaks were also observed for minor six proton signals as
highlighted with the arrow although some of them were over-
lapped each other or with major peaks. Taking the comparable
integral ratio of six proton signals into consideration, these minor
peaks are most likely derived from the anti-stereoisomer and the
syn/anti ratio was calculated as 60/40 (Fig. 3). This fact is in sharp
contrast to the diastereoselective cyclic trimerization of the bromo
counterpart (syn/anti = 100/0).23 We have preliminarily performed
the DFT calculations for DiMeO_C3A and DiBr_C3A. The geometry
optimization was executed using the B3LYP/6-31G(d) level of
theory. For both compounds, the syn-stereoisomer exhibited a
lower energy compared to the anti-stereoisomer due to the steric
repulsion of substituents (Fig. S3, ESI†). The energy difference
between syn and anti was somewhat larger for DiMeO_C3A
contrary to our expectations. Because the separation of syn- and
anti-stereoisomers was difficult by the chromatographic technique
and DiMeO_C3A was relatively unstable under the ambient
conditions, we gave up carrying out further characterization and
experiments.
In the second part of this paper, the stereoselectivity in the
dehydrative cyclic trimerization of 3-bromo-4-octylaminobenzoic
acid (6) was investigated (Scheme 2). The monomer was synthe-
sized in two steps starting from commercially available 4-amino-
benzoic acid (Scheme S2, ESI†). The cyclization of 6 was carried
out using SiCl4 and PPh3/Cl3CCCl3 as a dehydrative condensation
reagent in pyridine at 120 1C under microwave heating for 24 h.
Although higher molecular weight oligomers are contaminants of
the target compound on using SiCl4 similar to the synthesis of
DiMeO_C3A, the selective cyclic trimerization proceeded by using
PPh3/Cl3CCCl3 (Fig. S4, ESI†).
Fig. 2 1D NMR spectra and NOESY charts of DiMeO_C3A in CDCl3.
m/z = 812.4438 agreed well with the theoretical value of the Na
adduct of DiMeO_C3A (m/z = 812.4464). No higher molecular
weight oligomer such as a cyclic tetramer (m/z = 1075.5986) was
observed at all. Although the chemical purity of DiMeO_C3A was
ensured, the 1H-NMR spectrum showed a complex peak splitting
at around 3.7 ppm and 6.5 ppm region, which can be assigned to
methoxy and aromatic protons, respectively (Fig. 2, top). In order
to contemplate the result, the NOESY measurements were per-
formed in CDCl3 at room temperature. As indicated in the bottom
left chart of Fig. 2, the cross-peak was observed between NCH2
protons at 3.3 ppm and 4.0 ppm (marked with circle), and the
integral ratio of these signals was 1 : 1. This signal separation
would be ascribed to the diastereotopic environment of methylene
protons due to the chirality of DiMeO_C3A as in the case of
bromo-substituted derivative (DiBr_C3A).23 On the other hand, a
cross-peak was clearly observed for methoxy and aromatic protons
Scheme 2 Cyclic trimerization of 3-bromo-4-octylaminobenzoic acid (6).
1190 | New J. Chem., 2021, 45, 1187--1193
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Thus 3Br_C3A could be easily isolated by SiO2 column
Paper
chromatography as a pale yellow gummy solid in 30% yield.
The structure of 3Br_C3A was absolutely supported by the HR
ESI-MS measurement, in which the observed peak pattern
agreed well with the simulated pattern (Fig. S5, ESI†). Again, in
the 1H-NMR spectrum of 3Br_C3A (Fig. 4, top), methylene proton
signals adjacent to the amide nitrogen were separately observed
at 3.2 ppm and 4.2 ppm, which is ascribed to the diastereotopic
environment originated from the chirality of 3Br_C3A.
The HSQC chart of the aromatic region suggested that there
are four kinds of doublet ( J = 1.8 Hz), double doublet ( J = 1.8 Hz
and 8.1 Hz), and doublet ( J = 8.1 Hz) peaks (Fig. 4, bottom).
These four kinds of signals were found to be derived from the
stereoisomers (syn and anti) having the same molecular weight
because the comparable diffusion constants (D B 5 ꢁ 10ꢀ10 m2 sꢀ1
)
were obtained within the experimental error in the DOSY
analysis of a set of aromatic protons at 7.64 (a), 7.61 (b),
7.47 (c), and 7.38 (d) ppm (Fig. 5).25 All coupling patterns of
aromatic proton signals could be resolved by the DQF-COSY
measurement (Fig. S6, ESI†), and the syn/anti ratio was calculated
1
Fig. 5 Results of DOSY analysis for four aromatic protons (see the
Experimental section for details).
as 25/75 from the integral ratio in the H-NMR spectrum (Fig. 4,
top inset and Fig. 6).26
Fig. 6 Population of syn/anti stereoisomers in 3Br_C3A.
Even when the dehydrative cyclic trimerization of 6 using
PPh3/Cl3CCCl3 was carried out at lower temperature (100 1C),
an almost similar syn/anti ratio was observed (syn/anti = 27/73,
Fig. S7, ESI†). 3Br_C3A obtained using SiCl4 also exhibited the
same syn/anti ratio (25/75, not shown here). Namely, as for
3Br_C3A, the anti-stereoisomer was the major product unlike in
the case of DiBr_C3A. This anti preference characteristic of
3Br_C3A was qualitatively supported by the DFT calculations
because the anti-stereoisomer has a lower energy than the
syn-stereoisomer (Fig. S8, ESI†). This result can be easily
rationalized by the steric hindrance between bromo groups.
Azumaya and co-workers have previously reported that the
similar dehydrative cyclic condensation of 3-decyloxy-4-
(40-methoxybenzyl)aminobenzoic acid using Ph3PCl2 in 1,1,2,
2,-tetrachloroethane at 120 1C gave a cyclic trimer with a syn/
anti ratio of 33/66.27 Accordingly, the population of syn/anti
stereoisomers is supposed to be determined by the existence of
substituents at the ortho-position relative to the amide nitrogen
and less influenced by the character of the substituent.
Fig. 4 (top) 1D NMR spectrum of 3Br_C3A in CDCl3. Inset shows the
expanded spectrum of the aromatic region. (bottom) HSQC chart of
3Br_C3A in CDCl3.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 1187--1193 | 1191

View Article Online
Paper
NJC
The population of stereoisomers of the cyclic trimer, for example
the syn/anti ratio, was influenced by the substituent on the
benzene ring as evidenced by the comprehensive NMR analyses
and preliminary theoretical calculations. We believe that our
present findings have a great value for constructing supra-
molecular and polymer systems by using obtained compounds
as new building blocks.
Scheme 3 Reduction of the amide carbonyl group in 3Br_C3A.
The population of stereoisomers was almost irrespective of the Conflicts of interest
solvent character used in the NMR measurement. For example,
the syn/anti ratio in DMSO-d6 was calculated as 30/70 although
There are no conflicts to declare.
the complete signal assignment was difficult (Fig. S9, ESI†).
Therefore the interconversion between syn- and anti-conformers
of 3Br_C3A is not allowed due to the steric hindrance of the
Acknowledgements
substituent on the benzene ring and the rigid amide linker A part of this work was financially supported by Izumi Science
(Fig. S10, ESI†), which was further demonstrated by the and Technology Foundation and Toyota Physical and Chemical
VT-NMR measurement. As indicated in the top view of Fig. S11 Research Institute. We thank Ms Amimoto, the Natural Science
(ESI†), no peak coalescence was observed although the apparent Center for Basic Research and Development (N-BARD), Hiroshima
down-field shifts were seen for some aromatic proton signals University for the measurement of HR ESI-MS.
upon heating the 1,1,2,2-tetrachloroethane-d2 solution of
3Br_C3A. These peak shifts might be stemmed from the inter-
action between 3Br_C3A and the solvent although further experi-
Notes and references
ments are required. The integral ratio did not change by varying
the temperature. As for methylene proton signals adjacent to the
amide nitrogen (Fig. S11, ESI,† bottom), a peak only at around
1 C. J. Pedersen, J. Am. Chem. Soc., 1967, 89, 2495–2496.
2 A. Ikeda and S. Shinkai, Chem. Rev., 1997, 97, 1713–1734.
3 I. Pochorovski and F. Diederich, Acc. Chem. Res., 2014, 47,
3 ppm exhibited a down-field shift, but two peaks were separately
2096–2105.
observed even at 100 1C. We finally carried out the reduction of
4 S. J. Barrow, S. Kasera, M. J. Rowland, J. del Barrio and
O. A. Scherman, Chem. Rev., 2015, 115, 12320–12406.
5 L. Isaacs, Acc. Chem. Res., 2014, 47, 2052–2062.
6 W.-F. Lai, A. L. Rogach and W.-T. Wong, Chem. Soc. Rev.,
2017, 46, 6379–6419.
7 G. Crini, Chem. Rev., 2014, 114, 10940–10975.
8 T. Kakuta, T. Yamagishi and T. Ogoshi, Acc. Chem. Res.,
2018, 51, 1656–1666.
the amide carbonyl group28 for obtaining the cyclic trimer
3Br_C3Ared, which has a methylene group between the benzene
ring and nitrogen atom (Scheme 3). Although the reduction by
using LiAlH4 unexpectedly resulted in the production of unchar-
acterizable compounds, the combination of LiAlH4 with AlCl3
succeeded in obtaining the target compound. Because the anti-
stereoisomer should have three sets of doublet proton signals
with J = 1.8 Hz in the equal integral ratio, the proton signal at
7.57 ppm can be assigned to the syn-stereoisomer (Fig. S12, ESI†).
Thus, from the integral ratio, the syn/anti ratio of
10 P. J. Cragg and K. Sharma, Chem. Soc. Rev., 2012, 41,
3Br_C3Ared was calculated as 15/85. Even though the skeletal
rigidity is relaxed after reducing the amide carbonyl group, the
interconversion between syn- and anti-conformers is difficult
9 N. L. Strutt, H. Zhang, S. T. Schneebeli and J. F. Stoddart,
Acc. Chem. Res., 2014, 47, 2631–2642.
597–607.
11 Z. Liu, S. K. M. Nalluri and J. F. Stoddart, Chem. Soc. Rev.,
2017, 46, 2459–2478.
due to the steric hindrance of the substituent on the benzene
¨
12 C. Hippius, F. Schlosser, M. O. Vysotsky, V. Bohmer and
ring (vide supra). It can be speculated that the sterically crowded
syn-stereoisomer of 3Br_C3Ared partially decomposes as
deduced from the presence of a peak at m/z = 296.1009 in the
HR ESI-MS (Fig. S13, ESI†).
F. Wu¨rthner, J. Am. Chem. Soc., 2006, 128, 3870–3871.
¨
13 C. Hippius, I. H. M. van Stokkum, M. Gsanger, M. M.
Groeneveld, R. M. Williams and F. Wu¨rthner, J. Phys.
Chem. C, 2008, 112, 2476–2486.
14 F. Imabeppu, K. Katagiri, H. Masu, T. Kato, M. Tominaga,
B. Therrien, H. Takayanagi, E. Kaji, K. Yamaguchi,
H. Kagechika and I. Azumaya, Tetrahedron Lett., 2006, 47,
413–416.
Conclusions
The design and synthesis of cyclic oligomers with well-defined
conformations and functional groups are of much importance 15 I. Azumaya, H. Kagechika, K. Yamaguchi and K. Shodo,
not only in pursuing the host–guest chemistry but also for Tetrahedron Lett., 1996, 37, 5003–5006.
developing supramolecular and polymer materials with smart 16 A. Itai, Y. Toriumi, N. Tomioka, H. Kagechika, I. Azumaya
functions. In this work, we have investigated the stereoselectivity and K. Shodo, Tetrahedron Lett., 1989, 30, 6177–6180.
in the dehydrative cyclic trimerization of 4-alkylaminobenzoic 17 R. Yamakado, S. Sugimoto, S. Matsuoka, M. Suzuki,
acids having two methoxy groups and one bromo group. Y. Funahashi and K. Takagi, Chem. Lett., 2012, 41, 249–251.
1192 | New J. Chem., 2021, 45, 1187--1193
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021

View Article Online
NJC
Paper
18 K. Takagi, S. Sugimoto, R. Yamakado and K. Nobuke, J. Org. 24 M. Schulze, B. Michen, A. Fink and A. F. M. Kilbinger,
Chem., 2011, 76, 2471–2478. Macromolecules, 2013, 46, 5520–5530.
19 R. Yamakado, S. Matsuoka, M. Suzuki, K. Takagi, K. Katagiri 25 If the benzene ring of cyclic oligomers can rotate at room
and I. Azumaya, Tetrahedron, 2013, 69, 1516–1520.
20 H. Kakuta, I. Azumaya, H. Masu, M. Matsumura,
K. Yamaguchi, H. Kagechika and A. Tanatani, Tetrahedron
Lett., 2010, 66, 8254–8260.
temperature, four kinds of doublet signals may arise from
cyclic oligomers having the different ring size. In order to
clarify the origin of peak separation, we have performed the
DOSY measurement.
21 R. Yamakado, K. Mikami, K. Takagi, I. Azumaya, S. Sugimoto, 26 K. Katagiri, T. Ikeda, A. Muranaka, M. Uchiyama, M. Tominaga
S. Matsuoka, M. Suzuki, K. Katagiri, M. Uchiyama and
A. Muranaka, Chem. – Eur. J., 2013, 19, 11853–11857.
22 L. Escobar, Y.-S. Li, Y. Cohen, Y. Yu, J. Rebek Jr. and
P. Ballester, Chem. – Eur. J., 2020, 26, 8220–8225.
and I. Azumaya, Tetrahedron: Asymmetry, 2009, 20, 2646–2650.
27 The complete assignment of proton signals for syn- and
anti-compound is unsuccessful at the present stage because
the optical isomers (enantiomers and diastereomers) can-
not be separated by the chromatographic technique.
23 R. Yamakado, S. Matsuoka, M. Suzuki, D. Takeuchi,
H. Masu, I. Azumaya and K. Takagi, Chem. Commun., 28 R. Yamakado, S. Matsuoka, M. Suzuki, D. Takeuchi, H. Masu,
2015, 51, 5710–5713. I. Azumaya and K. Takagi, RSC Adv., 2014, 4, 6752–6760.
This journal is ©The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2021
New J. Chem., 2021, 45, 1187--1193 | 1193