HAL Id: cea-03250187
https://cea.hal.science/cea-03250187
Submitted on 4 Jun 2021
HAL is a multi-disciplinary open access
archive for the deposit and dissemination of sci-
entic research documents, whether they are pub-
lished or not. The documents may come from
teaching and research institutions in France or
abroad, or from public or private research centers.
L’archive ouverte pluridisciplinaire HAL, est
destinée au dépôt et à la diusion de documents
scientiques de niveau recherche, publiés ou non,
émanant des établissements d’enseignement et de
recherche français ou étrangers, des laboratoires
publics ou privés.
Are the physical properties of Xe@cryptophane
complexes easily predictable? the case of syn and anti
tris-aza-cryptophanes
Martin Doll, Patrick Berthault, Estelle Léonce, Céline Boutin, Thierry
Bueteau, Nicolas Daugey, Nicolas Vanthuyne, Marion Jean, Thierry Brotin,
Nicolas de Rycke
To cite this version:
Martin Doll, Patrick Berthault, Estelle Léonce, Céline Boutin, Thierry Bueteau, et al.. Are the
physical properties of Xe@cryptophane complexes easily predictable? the case of syn and anti tris-aza-
cryptophanes. Journal of Organic Chemistry, 2021, 86 (11), pp.7648-7658. �10.1021/acs.joc.1c00701�.
�cea-03250187�
1
Are the physical properties of Xe@Cryptophane complexes easily
predictable? The case of the syn and anti tris-aza-cryptophanes
Martin Doll,
†
Patrick Berthault,
‡
Estelle Léonce,
‡
Céline Boutin,
‡
Thierry Buffeteau,
§
Nicolas Daugey,
§
Nicolas Vanthuyne,
&
Marion Jean,
&
Thierry Brotin,*
, †
, Nicolas De Rycke,*
,†
†
Laboratoire de Chimie de l’ENS Lyon, (UMR 5182 CNRS-ENS-Université), Université Claude Bernard Lyon
1, F69342, Lyon, France
‡
Université Paris-Saclay, CNRS, CEA, Nanosciences et Innovation pour les Matériaux, la Biomédecine et
l'Energie (UMR 3685 CEA-CNRS), 91191, Gif-sur-Yvette, France
§
Institut des Sciences Moléculaires (UMR 5255 - Université-CNRS), Université de Bordeaux, 351 Cours de la
Libération, 33405 Talence, France
&
Aix Marseille Universit, Centrale Marseille, CNRS, iSm2 UMR 7313, 13397 Marseille, France
ABSTRACT: We report the synthesis and the optical resolution of C
3
-symmetrical tris-aza-
cryptophanes anti-3 and syn-4, as well as the study of their interaction with xenon via
hyperpolarized
129
Xe NMR. These molecular cages are close structural analogs of the two well-
known cryptophane-A (1; chiral) and cryptophane-B (2; achiral) diastereomers since these new
2
compounds differ only by the presence of three nitrogen atoms grafted on the same
cyclotribenzylene unit. The assignment of their relative (syn vs anti) and absolute
configurations was made possible thanks to the combined use of quantum calculation at the
DFT level and vibrational circular dichroism (VCD). More importantly, our results show that
despite the large structural similarities with cryptophane-A (1) and -B (2), these two new
compounds show very different behavior in the presence of xenon in organic solution. These
results demonstrate that a prediction of the physical properties of the xenon@cryptophane
complexes, only based on structural parameters, remains extremely difficult.
INTRODUCTION
Supramolecular structures based on poly-aromatic systems have attracted a lot of
attention in the past and this area of research is still very active today. For instance, the
pioneering work of Cram and co-workers in the 70-80’s has been a source of inspiration to
design new original structures capable of establishing very selective interaction with various
substrates in solution.
1
In the same period, other organic systems such as cryptophane
derivatives have emerged and this class of compound has been the subject of numerous studies
due to their ability to reversibly encapsulate atoms or small molecules in solution.
2
Thus,
cryptophane derivatives, which are hollow molecules with a characteristic globular shape, are
among the first artificial organic systems able to isolate a guest molecule from the bulk. This
property makes the identification of these complexes through NMR easier since the
encapsulated guest molecule usually shows a specific spectral signature, which is different from
the guest molecule in solution. Famous examples have been reported in the past with methane,
xenon, or even cationic species as guests in which nuclei of the encapsulated species show large
chemical shift difference with respect to the free guest present in the bulk.
3
3
Despite the great interest for these supramolecular systems, cryptophane derivatives
remain difficult to prepare and the production of novel structures remains limited even though
sustained efforts have been made by our group and others to develop the chemistry of these
molecules. Several issues related to the synthesis of these derivatives strongly limit the
possibility to prepare easily new cryptophane derivatives. For example, ring-closure reactions
leading to the formation of the two cyclotribenzylene units (CTB) are the main limitation, as
the electron-donating substituents must be grafted onto the benzene rings for the reaction to be
successful. Thus, for this reason the overwhelming majority of the cryptophane derivatives
reported in the literature has been prepared from vanillyl alcohol as a starting material, a
molecule that contains two oxygen atoms grafted on the benzene ring.
4
Despite the difficulties
in preparing novel cryptophane structures decorated with heteroatoms, the synthesis of sulfur-
containing cryptophanes has been reported by Hardie et al and Chambron et al via the direct-
coupling method.
5
Interestingly, this approach does not require the formation of a CTB ring at
a later stage of the synthesis but it cannot be easily extended to the synthesis of cryptophane
derivatives bearing heteroatoms other than sulfur. Of course, the synthesis of heteroatom-
decorated cryptophane derivatives can be considered, but their synthesis would require
additional chemical transformations once the cryptophane backbone is formed, which remains
a difficult task.
The synthesis of aza-congeners of cryptophane-A and cryptophane-B has not been
reported yet. This may appear surprising since when grafted onto a benzene ring, nitrogen is a
strong electron-donating atom and the formation of nitrogen-containing CTB derivatives has
been reported with excellent yields.
6
Nevertheless, the replacement of one or several oxygen
atoms present in the structures of cryptophane-A (1; chiral) or cryptophane-B (2; achiral) is not
straightforward and it represents an interesting synthetic challenge. In this article, we focus
exclusively on the synthesis of aza-cryptophane anti-3 and syn-4 bearing three nitrogen atoms
4
grafted on the same CTB cap (Chart 1). It is noteworthy that these new aza-derivatives anti-3
and syn-4 are both chiral compounds.
Cryptophane-A (1) and cryptophane-B (2) are diastereomers with small internal cavities
(V
vdw
= 90-100 Å
3
) and they differ only by the arrangement of the three linkers (anti
conformation for cryptophane-A and syn conformation for cryptophane-B).
7,8
Another way to
describe these two systems is to assign for each CTB unit a stereo-descriptor M or P.
9
Thus, for
the two structures reported in Chart 1, we can associate the PP absolute configuration (AC) for
the cryptophane-A derivative whereas cryptophane-B possesses the PM absolute configuration.
It is noteworthy that anti-cryptophane-A (1) and anti-3 are not defined by the same stereo-
descriptors (the same is true for syn-cryptophane-B (2) and syn-4). Indeed, the readers have to
keep in mind that the stereo-descriptors used for compounds anti-3 and syn-4 are different from
those used for the two cryptophane congeners A (1) and B (2) (Chart 1) due to the change of
priority between the oxygen and nitrogen atoms.
Replacement of three oxygen atoms by nitrogen atoms is expected to modify both the
electronic and magnetic properties of the cavity without affecting significantly the cavity size.
Thus, even though compounds 3 and 4 are expected to be very similar in shape to compounds
1 and 2, respectively, their molecular recognition properties are also expected to differ to some
extent. Moreover, the introduction of nitrogen atoms allows us to investigate the binding
properties of these hosts under different conditions. Indeed, protonation of the nitrogen atoms
under acidic condition should affect significantly the properties of these two hosts.
5
Chart 1: Chemical structures of cryptophane-A (1, chiral, D
3
symmetry), cryptophane-B (2; achiral, C
3h
symmetry) and aza-cryptophane congeners (3 and 4; both chiral, C
3
symmetry). Only a single enantiomer is shown
for chiral compounds, PP-1, PM-2, PM-3 and MM-4.
We report in this article the synthesis and the characterization of the two aza-
cryptophane anti-3 and syn-4. The two enantiomers of compounds anti-3 and syn-4 were first
separated by High-Performance Liquid Chromatography (HPLC) using chiral stationary
phases. Then, the chiroptical properties of the optically active compounds anti-3 and syn-4 were
investigated by Electronic Circular Dichroism (ECD) and Vibrational Circular Dichroism
(VCD) in order to determine their relative and absolute configuration. Finally, the formation of
their complexes with xenon in organic solution was reported under different conditions and
these results were compared with those previously obtained for cryptophane-A (1) and
cryptophane-B (2) in the same conditions.
RESULTS
Synthesis of aza-cryptophanes anti-3 and syn-4. The strategy used to prepare
compounds 3 and 4 relies on the so-called template method, which allows the formation of the
6
two CTB caps of the cryptophane derivatives at different stages of synthesis.
2b
First, the
preparation of a CTB decorated with three nitrogen atoms was performed using 4-amino-3-
methoxybenzyl alcohol (5) as starting material, according to a known procedure (Scheme 1,
route A).
10
This synthesis leads to the triamino-CTB 6 in three steps (acylation with acetic
anhydride, cyclization and deprotection of the acetyl moieties), with a 62% overall yield.
Finally, compound 6 was allowed to react with bromoacetyl bromide in the presence of DIPEA
to give compound 7 in 55% yield. Thus, using route A compound 7 was obtained with a 34%
overall yield but we noticed that the formation of compound 7 could be improved by using a
two steps approach (Scheme 1, route B). In this novel approach, compound 5 was allowed to
react with two equivalents of bromoacetyl bromide in presence of DIPEA to give compound 8
in 70% yield. In the presence of a HClO
4
/AcOH mixture, compound 8 gave rise to compound
7 in 86% yield, leading to a shorter synthesis and a better overall yield (60%).
Scheme 1: Synthesis of tris-aza-CTB 7 according to route A (four steps) and route B (two steps).
Thanks to the presence of three terminal bromomethyl functions, compound 7 can be
used as an interesting chemical platform to introduce new functionalities via S
N
2 reactions.
Thus, compound 7 was allowed to react with three equivalents of protected vanillyl alcohol 9
in the presence of K
2
CO
3
to give compound 10 in moderate yield (52%) after purification
7
(Scheme 2). Then, transformation of the three aromatic amide functions of compound 10 into
secondary aromatic amines was performed in THF, under reflux conditions, by using borane
dimethyl sulfide complex as the reducing agent. Purification on silica gel gave rise compound
11 in moderate yield (38%). Finally, the second ring closure reaction was performed by using
a mixture of perchloric acid and acetic acid at room temperature. Purification of the crude
product on silica gel followed by recrystallization in CH
2
Cl
2
/EtOH allowed us to isolate two
compounds. The first eluted compound was isolated with a yield of 1.2% and the second was
isolated with a 6% yield. The analysis of their respective
1
H NMR spectra revealed proton
signals characteristic of a cryptophane skeleton. These two
1
H NMR spectra show a lot of
similarities suggesting that both compounds are syn and anti-cryptophane diastereomers.
Indeed, the presence of four signals in the aromatic region of the
1
H NMR spectra allowed us
to confirm that these two compounds have C
3
-symmetry, as expected for anti-3 and syn-4.
Thanks to
1
H,
13
C and 2D-NMR (COSY, HSQC and HMBC) spectroscopy, the complete
assignment of the
1
H signals of these two derivatives was carried out (Figures S1-S18).
However, the relative configuration (syn vs anti) of the first and second eluted compounds is
not known at this stage and additional characterizations are necessary to determine this point.
In addition, compounds anti-3 and syn-4 both possess an inherent chiral structure due to their
lack of symmetry (C
3
group). Thus, the combined use of vibrational circular dichroism (VCD)
spectroscopy and density functional theory (DFT) calculations appeared here as the method of
choice for the attribution of their relative and absolute configuration. Indeed, even though the
two compounds show very similar structures, anti-3 and syn-4 should exhibit different VCD
spectra, as already observed with optically active cryptophane-A congeners decorated with nine
methoxy substituents.
11
8
Scheme 2: Synthesis of aza-cryptophanes anti-3 and syn-4.
Assignment of the relative and absolute configuration of 3 and 4. The two
enantiomers of the first and second eluted cryptophanes on silica gel have been separated by
HPLC using chiral stationary phase (Figures S19-S20). The two enantiomers of the first isolated
cryptophane derivative on silica gel have been separated on Chiralpak IE chiral column. Each
enantiomer of this derivative has been isolated with an enantiomeric excess (ee) >98%.
Similarly, the two enantiomers of second eluted cryptophane on silica gel have been separated
on the same chiral column but using a different mobile phase. These two enantiomers have also
been isolated with excellent ee (>99.5%). In the two separations, sufficient material was isolated
to study their chiroptical properties by VCD and ECD spectroscopy in order to determine their
relative and absolute configuration. The
1
H NMR spectra of the enantiopure cryptophanes (+)-
3, (–)-3, (+)-4 and (-)-4 are reported in Supporting Information (Figures S21-S24).
Specific Optical Rotation (SOR) values for the two enantiomers of compounds 3 and 4
have been measured in CH
2
Cl
2
and compared to the SOR values of compound 1 (Figure S25).
The first eluted enantiomer of 3, [CD()
254
]-3, exhibits negative values of SOR at 436, 546, 577
9
and 589 nm. Thus, the sign of the SOR value at 589 nm and the sign of the CD at 254 nm is the
same, which facilitates its identification. The SOR values of [CD()
254
]-3 are very close than
those measured for [CD()
254
]-1. The first eluted enantiomer of 4, [CD()
254
]-4, shows also
negative values of SOR at 436, 546, 577 and 589 nm, but these values are significantly lower
than those measured for [CD()
254
]-3.
The two enantiomers of the two compounds 3 and 4 were studied by IR and VCD
spectroscopy. The IR spectra recorded in CDCl
3
of the first and second eluted compounds
(silica gel) exhibit similar spectra in the spectral range 3600-950 cm
-1
, and only small
differences in the intensity of the IR bands are observed (Figures S26). Thus, it is not possible
to assign the relative configuration (syn vs anti) of the two compounds from IR spectra. The
VCD spectra recorded in CDCl
3
for the two enantiomers of compounds 3 and 4 are presented
in Supporting Information (Figures S27-S28). For each compound, the VCD spectra of the two
enantiomers are perfect mirror images, as expected for enantiopure compounds. Interestingly,
a comparison of the VCD spectra of the two compounds reveals large spectral differences in
the 1700-950 cm
-1
region (Figure 1). For instance, the
8b
C=C stretching vibration around 1610
cm
-1
exhibits a negative band for the first eluted compound on silica gel and the second eluted
enantiomer on Chiralpak IE column (Figure 1a, bottom), whereas the same stretching vibration
shows a positive-negative bisignate band for the second eluted compound on silica gel and the
first eluted enantiomer on Chiralpak IE column (Figure 1b, bottom). Likewise, the
19b
C=C
stretching vibration around 1510 cm
-1
exhibit a negative-positive bisignate band in Figure 1a
and a negative-positive-negative pattern in Figure 1b. More importantly are the spectral
differences in the 1300-1200 cm
-1
region. Ab-initio calculations of the VCD spectra at the DFT
level (B3LYP/6.31G**) for the PM-anti-3 and the PP-syn-4 molecules allow us to reproduce
with a good degree of confidence the experimental VCD spectra of the two compounds
considered in Figure 1a and 1b, respectively, allowing the determination of their relative and
10
absolute configurations.
12
It can be concluded that the first eluted compound on silica gel
corresponds to the anti-3 derivative. In turn, the second eluted compound on silica gel can
undoubtedly be attributed to the syn-4 derivative. In addition, the comparison between the
theoretical and experimental VCD spectra reveals that the PM (or MP) absolute configuration
can be attributed to the second (or first) eluted enantiomer of anti-3 on the Chiralpak IE
column. For the derivative syn-4, the VCD experiments allow us to assign the PP (or MM)
absolute configuration for the first (or second) eluted enantiomer on the chiral HPLC column.
The chemical structures of these four compounds are reported in Supporting Information
(Figure S29).
Figure 1: Experimental VCD spectra (bottom) of the first eluted compound on silica gel and the second eluted
enantiomer on Chiralpak IE column (a) and of the second eluted compound on silica gel and the first eluted
enantiomer on Chiralpak IE column (b) recorded in CDCl
3
at 25°C. Theoretical VCD spectra (top) for the PP-
anti-3 (a) and the PP-syn-4 (b).
As compounds PP-1 and PM-3 show large structural similarities, it is interesting to
compare their IR and VCD spectra in order to evaluate how the introduction of heteroatoms in
the cryptophane-A skeleton can affect the vibrational spectra. Of course, this comparison cannot
first eluted compound on silica gel & second eluted enantiomer
DFT: PM-anti-3
1700
1600
1500
1400
1300 1200 1100 1000
Wavenumbers, cm
-1
De (L mol
-1
cm
-1
)
-0.2
-0.0
0.2
0.4
0.6
0.8
a
-0.4
DFT: PP-syn-4
second eluted compound on silica gel & first eluted enantiomer
1700
1600
1500
1400
1300 1200 1100 1000
Wavenumbers, cm
-1
De (L mol
-1
cm
-1
)
-0.2
-0.0
0.2
0.4
0.6
0.8
-0.4
b
11
be done with compounds syn-4 and cryptophane-B (2) since compound 2 is not chiral. It can be
noticed that the IR spectra of compounds anti-1 and anti-3 are very similar and the main spectral
differences are observed below 1250 cm
-1
(Figure S30). On the other hand, more important
spectral differences are observed in their VCD spectra (Figure S31). Indeed, the overall
intensity of the VCD spectrum of compound PM-3 is measured with a significant lower
intensity than its parent molecule PP-1, probably as a consequence of the breakdown of its
molecular symmetry. Moreover, this spectrum reveals large spectral differences with respect to
the VCD spectrum of PP-1 in the 1400-1350 and 1250-1200 cm
-1
region related to the linkers
and associated with the wagging CH
2
chains and
a
C-O-C stretching mode, respectively.
The ECD spectra of compounds have been recorded in CH
2
Cl
2
and THF for compound
anti-3 and in CH
2
Cl
2
, CHCl
3
, CH
3
CN and THF for compound syn-4 (Figures S32-S37). In
CH
3
CN the MM-syn-4 enantiomer shows several positive-negative bands of different
intensities between 220 and 340 nm (Figure S36). For instance, this enantiomer shows two
negative Cotton bands of relatively high intensities at 292 nm (L mol
-1
cm
-1
) and
250 nm (L mol
-1
cm
-1
). Two positive Cotton bands of similar intensities are also
observed for this enantiomer at 269.5 nm (L mol
-1
cm
-1
) and at 232 nm (
L mol
-1
cm
-1
). In addition, a close examination of the ECD spectrum reveals the presence of
additional Cotton bands of very week intensities at higher wavelengths. Indeed, the same
enantiomer shows one negative Cotton band at 318 nm ( L mol
-1
cm
-1
) and two
positive Cotton bands at 328 nm and 306 nm ( L mol
-1
cm
-1
). It is noteworthy that the
PP-syn-4 enantiomer shows a perfect mirror image in the same condition. Changing the nature
of the solvent has little impact of the global shape of the ECD spectra except for the weak
Cotton bands located between 300 and 340 nm.
12
Additional ECD experiments were performed in CH
2
Cl
2
with the two enantiomers of
syn-4 in the presence of various amount of trifluoroacetic acid (TFA). For instance, compound
MM-syn-4 shows large spectral differences as the amount of TFA added to the CH
2
Cl
2
solution
increases. These spectral differences result in a significant decrease of all the Cotton bands
observed between 230 and 300 nm (Figure S38a). The PP-syn-4 enantiomer behaves similarly
(Figure S38b). Under similar experimental conditions, we observed that the two enantiomers
of anti-3 are also affected by the addition of TFA into the solution (Figures S39 a,b). The
corresponding spectra of anisotropy factors (g = ε/ε) were also calculated and are reported in
Supporting Information (Figures S32-S39).
Study of the interaction of anti-3 and syn-4 with xenon using hyperpolarized
129
Xe
NMR. Laser-polarized xenon prepared in the batch mode has been introduced in C
2
D
2
Cl
4
solutions containing compounds anti-3 and syn-4 and studied at 11.7 T and 293 K. The
129
Xe
spectrum of xenon in the anti-3 solution reveals two sharp signals that can be easily identified.
The intense signal calibrated at 223 ppm (Figure 2a) corresponds to xenon dissolved in
tetrachloroethane-d
2
. The second signal, shifted toward low frequencies (ppm),
corresponds to xenon caged in the compound anti-3. Under the same experimental conditions,
the
129
Xe NMR spectrum of syn-4 is characterized by two sharp signals indicating that here also
xenon experiences a slow in-out exchange dynamics with the compound syn-4 at 293 K and
11.7 T (Figure 2b). In addition to the strong
129
Xe signal corresponding to xenon free in 1,1,2,2-
tetrachloroethane-d
2
, calibrated at 223 ppm, another sharp signal shifted towards low
frequencies and located at ppm corresponds to xenon caged in syn-4.
13
Figure 2: One-scan
129
Xe NMR spectra of (a) Xe@anti-3 (c = 3.0 mM) and (b) Xe@syn-4 (c = 3.3 mM) recorded
in C
2
D
2
Cl
4
at 293 K and 11.7 T.
Since compounds anti-3 and syn-4 belong to the families of cryptophane-A (1) and B
(2) respectively, it seems interesting to compare their
129
Xe NMR spectra in the same
experimental conditions. The most significant difference is between the complexes with syn-2
(cryptophane-B) and syn-4, the former giving rise to a fast in-out xenon exchange situation at
11.7 T and 293 K. Consequently, at this temperature its
129
Xe NMR spectrum does not exhibit
a specific resonance frequency for the complex Xe@2.
8
Conversely, the
129
Xe NMR spectrum
of a solution containing a mixture of cryptophanes anti-3 and anti-1 reveals three signals
(Figure 3), indicating slow in-out xenon exchange for both complexes. The two high-field
shifted
129
Xe NMR signals, at 65.9 and 51.9 ppm correspond to xenon encapsulated in
compounds 1 and 3, respectively. From the linewidths of the caged xenon signals, a clear
difference in the in-out exchange dynamics of xenon can be observed between the two
cryptophanes. This exchange is faster for the system Xe@1 than for Xe@3. At first sight this
may appear surprising since these compounds are structurally identical (only three oxygen
atoms of 1 have been replaced by three nitrogen atoms) and possess similar cavity size. From
this experiment it can be concluded that the presence of nitrogen atoms in the skeleton of the
compounds anti-3 and syn-4 reduces significantly the xenon in-out exchange rate compared to
what is observed with the corresponding cryptophanes containing ethylenedioxy linkers (anti-
14
1 and syn-2, respectively). The knowledge of the association constant for the Xe@1 complex
(3900 M
-1
), allow us to estimate the association constant K
a
= 2300 M
-1
for the Xe@3 complex.
From a separate
129
Xe NMR experiment performed on the mixture of cryptophanes 3 and 4, the
association constant with 4 could be estimated to be 2800 M
-1
. Keep in mind that a large
uncertainty is encountered with these successive experiments (cumulative error).
Figure 3: One-scan
129
Xe NMR spectrum of compound 3 (c =1.6 mM) and cryptophane-A (1) (c = 1.3 mM)
recorded at 293 K in C
2
D
2
Cl
4
.
Finally, we have studied the effect of protonation of the three aromatic amine moieties
present in the cryptophane skeleton of anti-3 and syn-4 on their
129
Xe NMR spectra. The
129
Xe
NMR spectra of Xe@anti-3 and Xe@syn-4 complexes have been recorded in C
2
D
2
Cl
4
solutions
at 293 K with an excess of TFA. It can be seen that addition of 1 µL TFA to these solutions has
a drastic effect on
129
Xe NMR spectra of both Xe@anti-3 and Xe@syn-4 complexes (Figures
S40-S41). For instance, with addition of TFA the Xe@syn-4 signal is low-field shifted to reach
a value of 99.7 ppm (δ = 27.6 ppm with respect to neutral Xe@syn-4 complex) and is
broadened. More surprisingly, in the same conditions the Xe@anti-3 complex gives rise to a
very broad signal that is difficult to distinguish from noise on the
129
Xe NMR spectrum at 293
K.
15
DISCUSSION
Anti-3 and syn-4 aza-cryptophanes have been obtained in low yield and they have been
more difficult to isolate than their congeners cryptophane-A and cryptophane-B due to the
presence of nitrogen atoms in the cryptophane skeletons. It is noteworthy that the formation of
anti-3 and syn-4 takes place during the same reaction whereas two different protocols are
necessary to isolate cryptophane-A and cryptophane-B. At first sight, the difficulties to obtain
compounds anti-3 and syn-4 may appear surprising since the synthetic pathway used to prepare
these two derivatives is similar to the one used for the preparation of compounds anti-1 and
syn-2. Indeed, the second ring closing reaction involves only benzenic groups decorated with
oxygen atoms and the CTB decorated with aromatic amine substituents does not directly
participate to the cyclization reaction. Thus, the large difference in the isolated yields for these
two compounds suggests that the protonation of the three nitrogen atoms has a strong impact
on the conformation of the linkers, which in turn hinder the second ring closing reaction.
Attempts have been made to improve the yield of the reaction by changing the experimental
conditions. For instance, the replacement of HClO
4
/AcOH by a HClO
4
/MeOH mixture at room
temperature also resulted in the formation of cryptophane anti-3 and syn-4 in small amount.
The formation of these two diastereomeric compounds was detected by
1
H NMR spectroscopy
but they were not isolated. A change of the nature of the acid did not allow us to improve the
yield of the reaction. For instance, the use of a HCOOH/CHCl
3
(v: 1/1) mixture at 60°C under
diluted conditions resulted in the formation of a complex mixture of compounds, which were
not characterized, and the expected syn and anti-cryptophanes were not detected from the crude
material. These conditions are those usually reported to produce in high yield the cryptophane-
A (1) derivative and its congeners.
2
Similarly, the use of milder conditions did not allow us to
16
improve the yield of compounds anti-3 and syn-4. For instance, the use of scandium triflate as
Lewis acid has been reported to promote the ring closing reaction of many different systems
such as cyclotriveratrylene, hemicryptophane and cryptophane derivatives.
13
Unfortunately, the
use of this reagent in CH
2
Cl
2
or in CH
3
CN did not allow us to promote the formation of the two
desired diastereomers. It should be noted that these experimental conditions were also used with
the cryptophane precursor 10, which can also be subjected to the Friedel-Crafts reaction under
acidic conditions. In all cases, a complex mixture of compounds was observed and the analysis
of the
1
H NMR spectrum of the crude mixture did not allow us to detect the desired compounds.
VCD spectroscopy associated with DFT calculations have been used to assign the
relative (anti vs syn) and absolute configurations of compounds 3 and 4. In contrast to
cryptophane-A and B, among which only cryptophane-A is chiral, both cryptophane derivatives
anti-3 and syn-4 are optically active and possess the same molecular C
3
-symmetry. As the
determination of the stereochemistry of each isolated cryptophane cannot be rapidly established
from the analysis of their respective
1
H NMR, chiroptical techniques have been privileged in
this article for the unambiguous determination of their stereochemistry. Previous studies
performed on highly substituted cryptophane derivatives with C
3
-symmetry have revealed that
VCD and ROA spectra of anti and syn diastereomers displayed large spectral differences.
11,14
In these examples, DFT calculations (B3PW91/6-31G**) reproduced fairly well the main
features of each spectrum, thus allowing to assign with a good degree of confidence the
stereochemistry of each compounds.
As expected, compounds anti-3 and syn-4 show large spectral differences and DFT
calculations (B3LYP/6-31G**) allow to reproduce fairly well these spectra. The main spectral
differences between anti-3 and syn-4 concerns the VCD bands related to the C=C stretching
vibrations (
8b
C=C and
19b
C=C around 1610 cm
-1
and 1510 cm
-1
, respectively) and those in
the 1300-1200 cm
-1
region related to alkyls chains and
a
C-O-C stretching vibration. Thus, this
17
study allows the assignment of the anti-diastereomer to the first eluted compound on silica gel.
In turn, the syn-diastereomer can be assigned to the second eluted compound. At first glance,
this result is surprising considering that in the large majority of examples the cryptophane
derivatives having the anti-stereochemistry are usually the major products. In this study, the
syn-diastereomer is formed preferentially (6%) compared to its anti-congener (1.2%) even
though the yield considered here are low. In addition, the comparison between the experimental
and theoretical VCD spectra allows the assignment of the AC of each enantiomer of the two
diastereomers. The PM (MP) absolute configuration has been attributed to the second (first)
eluted enantiomer of anti-3, whereas the PP (MM) absolute configuration is related to the first
(second) eluted enantiomer of syn-4.
Figure 4: ECD spectra of anti-PM-3 (red spectrum) and anti-PP-1 (black spectrum;
cryptophane-A) derivatives recorded in THF at 298 K.
The ECD spectra of compound anti-PM-3 reveal several spectral differences with
respect the parent anti-PP-1 derivative (Figure 4). These spectral differences come from the
fact that three oxygen atoms of compound anti-1 have been replaced by three nitrogen atoms.
Indeed, this chemical modification has two important consequences. First, it leads to a
18
breakdown of the molecular symmetry with respect to the parent D
3
-symmetric molecule 1. The
anti-3 compound has C
3
symmetry and due to the excitonic coupling model, each Cotton band
present on the ECD spectrum is the combination of several excited states whose number,
position and intensity (defined by their rotational strength) are different from those observed
with the anti-1 compound having D
3
symmetry.
15
Second, the nitrogen atom attached to a
benzene ring has a stronger electron-donating effect than an oxygen atom and, as a
consequence, it causes a larger rotation of the electronic transitions moment for the excited
states
1
L
b
,
1
L
a
and
1
B
b
. Finally, in addition to these electronic effects, it cannot be excluded that
the anti-3 compound adopts different linker conformations compared to what is observed with
cryptophane-A. These conformational changes may modify the shape of their ECD spectra.
The most interesting characteristics of anti-3 and syn-4 compounds derive from their
molecular recognition properties with xenon. Indeed, these two compounds show behaviors
very different from those observed with the parent molecules 1 and 2. At first glance, these
differences are difficult to interpret because compounds 1 and 3 on the one hand and,
compounds 2 and 4 on the other hand, share great structural similarities. For instance, replacing
C-O bonds by C-N bonds is not expected to change significantly the cavity size of compounds
3 and 4 compared to their parent derivatives. Our results reveal that xenon in the Xe@3 and
Xe@4 complexes resonates in the same frequency region than the Xe@1 complex. This is
expected because the encapsulated xenon in compounds 1-4 is surrounded by six benzene
groups decorated with electron-donating atoms (N, O for 3-4 and O, O for 1-2). However,
observation of the
129
Xe NMR spectra reveals that the replacement of three oxygen atoms by
three nitrogen atoms has a strong effect on the in-out xenon exchange rate. In order to explain
the large differences in behavior observed between hosts 3-4 and derivatives 1-2, we assume
that the aromatic amine groups can easily establish a hydrogen-bonding network with the
residual water molecules. These water molecules may be present in compounds 3 and 4 before
19
doing the
129
Xe NMR experiments or they may come from the solvent C
2
D
2
Cl
4
. In such an
apolar solvent, the aromatic amine groups can favor the formation of a water cluster near the
portal of cryptophanes 3-4. It is noteworthy that the presence of water molecule within the
cavity of cryptophane-A had already been identified in the solid state, however such
encapsulation phenomenon could not be detected here with hosts 3-4 in C
2
D
2
Cl
4
solutions.
16
In
this respect,
1
H NMR and
1
H-
1
H NOESY spectra of host 4 recorded in C
2
D
2
Cl
4
both reveal a
slow exchange between the free water signal at 1.6 ppm and a signal at 5.0 ppm which is
assigned to clusters of water close to the cryptophane portals. (Figures S42-S43). Therefore,
water molecules are likely to build a strong network of hydrogen bonds around the NH groups,
which in turn can impact the in-out exchange dynamics of xenon. This result, which has not
been observed with other cryptophane to date, sheds light on the possible role of water in the
overall properties of the xenon-cryptophane complexes.
Herein, the most striking result comes from the analysis of the
129
Xe NMR spectrum of
syn-4 that shows a very sharp spectral signature whereas no signal corresponding to the Xe@2
complex could be detected under the same experimental conditions (fast in-out exchange at the
NMR time scale). Surprisingly, it can be noticed that the Xe@3 and Xe@4 complexes share
more similarities with the Xe@cryptophane-111 complex that possesses a smaller inner cavity
(V = 81 Å
3
) and a large affinity for xenon in the same conditions.
17
It is noteworthy that upon acidification with trifluoroacetic acid a drastic change of the
129
Xe NMR spectra is observed. Indeed, upon acidification the electron donor properties of the
aromatic amine groups can be easily reversed and transformed into a strong electron attractor.
This should interfere with xenon complexation by cryptophanes 3-4. The acidification of the
medium has a strong effect on the shape of the signal of the encapsulated xenon and it results
in an increase of the exchange dynamics of xenon for the two complexes Xe@anti-3 and
Xe@syn-4. Even though, this effect is difficult to explain in details it is reasonable to assume
20
that the protonation of the three aromatic amine moieties has an effect on the conformation of
the three linkers. Such hypothesis is also supported by the strong modifications observed for
ECD spectra of hosts 3-4 upon acidification of the bulk which reveals potential significant
conformational changes. Protonation of the nitrogen atoms may also destroy the water cluster
present around hosts 3-4. Both parameters may affect drastically the in-out exchange dynamics
of xenon.
The large difference in the behavior of the cryptophane-A and cryptophane-B
congeners in the presence of xenon rise some comments. The two examples cited in this article
reveal that very slight change in the chemical structure of the cryptophane skeleton can have a
tremendous effect on the overall physical properties of the Xe@cryptophane complexes. Thus,
the study of cryptophanes 3 and 4 in the presence of xenon shows that the physical properties
of the Xe@cryptophane complexes are difficult to predict accurately on the sole basis of the
chemical structure of the studied cryptophane.
These results may have important consequences in the design of cryptophane biosensors
for
129
Xe MRI applications. Pines and co-workers have shown that encapsulated xenon in
cryptophane cavities could be used as a molecular tracer for the detection of biological events
at the molecular level using hyperpolarized
129
Xe NMR.
18
Since then, many different molecular
systems have been developed by our group and others to detect biomolecules or analytes by
this means.
19
So far, the overwhelming majority of organic systems cited in the literature use
the cryptophane-A skeleton to build up these biosensors.
20
A modification of the cryptophane
backbone structure to improve the characteristics of these biosensors is desired to improve
solubility in physiological media, the xenon binding constant or more importantly the xenon in-
out exchange dynamics that plays a key role in the detection of these molecular tracers.
21
This
work demonstrates that the main characteristics of the Xe@cryptophane complexes cannot be
easily predicted based on structural parameters. Therefore, this may complicate the design of
21
new molecular tracers optimized for MRI application. On the other hand, quantum calculations
can be used to predict some characteristic of these supramolecular complexes. For instance, this
approach has been successfully used to predict precisely the chemical shift of the encapsulated
xenon in cryptophane-111 derivatives decorated with a different number of hydroxyl
functions.
22
However, a prediction of the physical properties of xenon present within
cryptophane cavities is a difficult and time demanding task. For instance, the accurate
prediction of
129
Xe properties inside molecular hosts requires the use of particular functionals
density methods to take into account the dispersion energy to describe the host-guest interaction
and relativistic effects have also to be included in the calculation in the case of xenon. In
addition, it is noteworthy that a prediction of the in-out xenon exchange dynamics cannot be
achieved by quantum calculations methods. The results reported in this article also suggest that
the solvent and especially water has to be taken into consideration in these calculations to
predict accurately the physical properties of these supramolecular systems.
CONCLUSION
We report the synthesis of two cryptophane-A and B congeners decorated with nitrogen
atoms. The introduction of these heteroatoms yields to a breakdown of the molecular symmetry
and, as a consequence, both compounds show an inherent chiral structure. The identification of
the relative and absolute configurations of compounds anti-3 and syn-4 has been established
thanks to the combined use of DFT calculations and VCD spectroscopy. As a result of the
replacement of three oxygen atoms in compound anti-1, the new chiral compound anti-3
exhibits very different chiroptical properties from those observed with cryptophane-A.
Interestingly, these new derivatives show remarkable features in organic solution in the
presence of hyperpolarized xenon and large differences are observed in their
129
Xe NMR
22
spectra compared to those obtained with the cryptophane-A (1) and cryptophane-B (2)
derivatives. For instance, both complexes Xe@3 and Xe@4 complexes exhibit very sharp
129
Xe
NMR signals, which are characteristic of a slow in-out exchange dynamics of xenon. On the
other hand, this exchange appears to be much faster in the case of cryptophane-A and
cryptophane-B than for compounds anti-3 and syn-4 whereas the structures of these compounds
are similar. Interestingly, we show that the in-out exchange dynamics of xenon of the Xe@3
and Xe@4 complexes can be modified upon acidification of the bulk by addition of TFA. The
new derivatives 3 and 4 share structural similarities with cryptophane A and B, respectively,
but their behavior in the presence of xenon reveals large differences, suggesting that a
prediction of the physical properties of these complexes remains difficult on the basis of
structural comparisons alone. In addition, the unusual characteristics observed with Xe@3 and
Xe@4 complexes suggest a possible role of water molecules in the behavior of these complexes.
Work is underway to understand how water molecules interact with cryptophane derivatives.
EXPERIMENTAL DETAILS
Mass spectra (HRMS) were performed by the Centre de Spectromtrie de Masse,
University of Lyon. Analyses were performed with a hybrid quadrupole-time-of-flight mass
spectrometer, microToF QII equipped with an electrospray ion source. Data Analysis 4.0 was
used for instrument control, data collection, and data treatment. HRMS analyses were
performed in full scan MS with a mass range from 50 to 2000 Da at an acquisition rate of 1 Hz.
Transfer parameters were as follows: RF Funnel 1, 200 V; RF Funnel 2, 200 V; hexapole, 50
V; transfer time, 70 μs; and PrePulse storage time, 1 μs. Before each acquisition batch, external
calibration of the instrument was performed with a sodium formate clusters solution.
1
H and
13
C NMR spectra were recorded at 300 or 400 and 75.5 or 100.6 MHz, respectively. Chemical
shifts are referenced to Me
4
Si (
1
H,
13
C). Structural assignments were made with additional
23
information from gCOSY, gHSQC, and gHMBC experiments. Column chromatographic
separations were carried out over Merck silica gel 60 (0.040−0.063 mm). Analytical thin-layer
chromatography (TLC) was performed on Merck silica gel TLC plates, F-254. The solvents
were distilled prior to use: DMF and CH
2
Cl
2
from CaH
2
, THF from Na/benzophenone, and
pyridine from KOH.
HPLC Separation. The two enantiomers of anti-3 were separated on semi-preparative
Chiralpak IE (250 x 10 mm) chromatographic column. A mixture of EtOH + Et
3
N (0.1%
v/v)/CH
2
Cl
2
(50/50) was used as a mobile phase (flow-rate = 5 mL/min). UV detection was
performed at 254 nm. The two enantiomers of anti-3 have been successfully separated with an
enantiomeric purity 98%. Similarly, the two enantiomers of syn-4 were separated on semi-
preparative Chiralpak IE (250 x 10 mm) chromatographic column. A mixture of Hexane/EtOH
+ Et
3
N (0.1% v/v)/CH
2
Cl
2
(10/40/50) was used as a mobile phase (flow-rate = 5 mL/min). UV
detection was performed at 254 nm. The two enantiomers of syn-4 have been successfully
separated with an enantiomeric purity 99.5%.
UV-visible and ECD spectroscopy. ECD spectra were recorded in CH
2
Cl
2
, CHCl
3
,
CH
3
CN and THF at 293 K. Quartz cell with a path length of 0.2 cm and 1 cm were used for the
ECD and UV-visible experiments. A concentration range of 8 × 10
-5
to 1 × 10
-4
moles per liter
were used for ECD and UV-visible experiments. The ECD spectra were recorded in the
wavelength range of 225 – 400 nm with a 0.5 nm increment and a 1 s integration time. The
spectra were processed with standard spectrometer software. A smoothing procedure was
applied by using a third‐order least‐square polynomial fit when necessary.
129
Xe NMR spectroscopy. Xenon enriched at 83% in isotope 129 was hyperpolarized
via Spin Exchange Optical Pumping in the batch mode, using our home-built setup described.
23
The cryptophanes were solubilized in tetrachloroethane-d
2
and placed in 5-mm NMR tubes
24
capped with J. Young’s valves. The transfer of hyperpolarized xenon into these tubes previously
evacuated was made through a vacuum line in the fringe field of the NMR magnet. All the
129
Xe
NMR experiments were run at 11.7 T and 293 K.
VCD Spectroscopy and DFT calculations. The IR and VCD spectra were recorded
with a FTIR spectrometer equipped with a VCD optical bench, following the experimental
procedure previously published.
24
Samples were held in a 250 m path length cell with BaF
2
windows. IR and VCD spectra of the two enantiomers of anti-3 and syn-4 were measured in
CDCl
3
at a concentration of 15 mM.
All DFT calculations were carried out with Gaussian 09.
25
Preliminary conformer
distribution search of anti-3 and syn-4 was performed at the molecular mechanics level of
theory, employing MMFF94 force fields incorporated in ComputeVOA software package.
Around thirty conformers within roughly 2 kcal/mol of the lowest energy conformer were kept
and further geometry optimized at the DFT level using B3LYP functional and 6-31G** basis
set. Finally, only the four lowest energetic geometries were kept to predict the IR and VCD
spectra of anti-3 and syn-4. The four selected conformers exhibited a gauche,gauche,gauche
conformations of the three linkers of anti-3 and syn-4. Vibrational frequencies, IR and VCD
intensities were calculated at the same level of theory. For comparison to experiment, the
calculated frequencies were scaled by 0.97 and the calculated intensities were converted to
Lorentzian bands with a full-width at half-maximum (FWHM) of 14 cm
-1
.
EXPERIMENTAL PROCEDURE
Precursor 8. A solution of amine 5 (1.00 g, 6.54 mmol) in dry THF (30 mL) was set
under an argon atmosphere and cooled to 0 °C with an ice bath. DIPEA (2.8 mL, 16.3 mmol)
and bromoacetyl bromide (1.4 mL, 16.3 mmol) were added dropwise to the solution, which was
25
stirred at room temperature overnight. The originally pale-yellow solution turned into a brown
suspension. Water (100 mL) was added to the mixture, which was extracted three times by
EtOAc (3 x 100 mL). The organic layers were dried over sodium sulfate and the solvent was
evaporated under reduced pressure. The crude was purified using column chromatography
(SiO
2
, eluent: CH
2
Cl
2
) to gain 8 as a beige powder (1.8 g, yield: 70%).
1
H NMR (300 MHz,
CDCl
3
, 25 °C): δ 8.80 (s, 1H), 8.32 (d, 1H, J = 8.2 Hz), 6.97 (dd, 1H, J = 1.6 Hz, J = 8.2 Hz),
6.92 (d, 1H, J = 1.6 Hz), 5.17 (s, 2H), 4.02 (s, 2H), 3.93 (s, 3H), 3.87 (s, 2H).
13
C{
1
H} NMR
(75 MHz, CDCl
3
, 25 °C): δ 167.1, 163.3, 148.3, 131.4, 127.2, 121.3, 119.5, 110.3, 67.8, 56.0,
29.6, 25.8. HRMS (ESI) m/z: [M+Na]
+
calcd for C
12
H
13
Br
2
NNaO
4
415.9104; found 415.9102.
Aza-CTB derivative 7. Perchloric acid (50 mL) was added dropwise at room
temperature to a solution of compound 8 (4.30 g, 10.8 mmol) dissolved in acetic acid (25 mL)
under argon atmosphere. A white precipitate appeared within 30 minutes after the addition. The
reaction mixture was stirred at room temperature for additional 16 hours. Then the mixture was
poured into water (200 mL). The solid was filtered and washed three times with water (3 x 20
mL), twice with ethanol (2 x 20 mL) and twice with Et
2
O (2 x 20 mL). After drying in vacuo,
compound 7 was collected as a white solid (2.37 g, 86%).
1
H NMR (300 MHz, DMSO-d
6
, 25
°C): δ 9.46 (s, 3H), 8.04 (s, 3H), 7.00 (s, 3H), 4.78 (d, 3H, J = 13.4 Hz), 3.83 (s, 9H), 3.56 (d,
3H, J = 13.3 Hz).
13
C{
1
H} NMR (75 MHz, DMSO-d
6
, 25 °C): δ 165.2 (3C), 148.7 (3C), 136.9
(3C), 131.6 (3C), 125.4 (3C), 123.7 (3C), 112.7 (3C), 56.1 (3C), 35.8 (3C), 30.8 (3C). HRMS
(ESI) m/z: [M+Na]
+
calcd for C
30
H
30
Br
3
N
3
NaO
6
787.9577; found 787.9541.
Aza-CTB derivative 10. Compound 7 (2.00 g, 2.60 mmol) and potassium carbonate
(4.0 g, 29 mmol) were suspended in dry DMF (50 mL). A solution of precursor 9 (2.4 g, 10
mmol) in dry DMF (10 mL) was added. The mixture was heated to 70 °C (oil bath) for 16 h.
After cooling down to room temperature, water (300 mL) was added to precipitate the product,
which was filtered, washed with water (3 x 40 mL) and Et
2
O (40 mL) and dried under reduced
26
pressure. Compound 10 was obtained as a poorly soluble slight brown powder (1.73 g, 54%).
1
H NMR (400 MHz, CDCl
3
, 25 °C): δ 9.09 (s, 3H), 8.43 (s, 3H), 7.01-6.84 (m, 12H), 4.78 (d,
3H, J = 13.6 Hz), 4.76-4.40 (m, 15H), 3.95-3.85 (m, 21H), 3.65 (d, 3H, J = 13.4 Hz), 3.54 (m,
3H), 1.95-1.45 (m, 18H).
13
C{
1
H} NMR (101 MHz, CDCl
3
, 25 °C): δ 166.4 (3C), 149.9 (3C),
147.3 (3C), 146.6 (3C), 135.6 (3C), 133.3 (3C), 131.2 (3C), 125.4 (3C), 121.0 (3C), 120.5 (3C),
115.3 (3C), 112.1 (3C), 111.7 (3C), 97.7 (3C), 69.8 (3C), 68.6 (3C), 62.3 (3C), 56.0 (3C), 55.9
(3C), 36.5 (3C), 30.6 (3C), 25.5 (3C), 19.5 (3C). HRMS (ESI) m/z: [M+Na]
+
calcd for
C
69
H
81
N
3
NaO
18
1262.5407; found 1262.5413.
Aza-CTB derivative 11. Triamide CTB 10 (4.51 g, 3.64 mmol) was suspended in THF
(140 mL) and the mixture was stirred under an argon atmosphere. The solution was cooled
down to 0 °C with an ice bath and borane dimethyl sulfide (15 mL of a 2 M solution in THF)
was added dropwise to the suspension. The mixture was then heated to 50 °C (oil bath) for 3
hours. The remaining borane was quenched with methanol and the solvents were evaporated
under reduced pressure. Water (200 mL) and DCM (200 mL) were added to the residue. The
aqueous layer was extracted twice with dichloromethane (2 x 100 mL). The combined organic
layers were dried on anhydrous sodium sulfate and the solvent was evaporated under reduced
pressure. The crude was purified using column chromatography (SiO
2
, eluent: EtOAc -
petroleum ether (8:2)) to give rise to compound 11 (1.65 g, 38%) as a glassy product.
1
H NMR
(400 MHz, CDCl
3
, 25 °C): δ 6.93-6.80 (m, 9H), 6.71 (s, 3H), 6.62 (s, 3H), 4.73 (d, 3H, J = 13.6
Hz), 4.71 (d, 6H, J = 11.7 Hz), 4.43 (d, 6H, J = 11.7 Hz), 4.18 (m, 6H), 3.95-3.86 (m, 3H), 3.85
(s, 9H), 3.69 (s, 9H), 3.57-3.78 (m, 9H), 3.47 (d, 3H, J = 13.8 Hz), 1.92-1.48 (m, 18H).
13
C{
1
H}
NMR (101 MHz, CDCl
3
, 25 °C): δ 149.6 (3C), 147.6 (3C), 145.8 (3C), 136.2 (3C), 132.4 (3C),
131.6 (3C), 128.1 (3C), 120.5 (3C), 113.9 (3C), 111.9 (3C), 111.3 (3C), 111.2 (3C), 97.5 (3C
– double signal assigned to diastereoisomers), 77.2 (3C), 68.6 (3C), 68.1 (3C), 62.2 (3C), 55.8
27
(3C), 55.5 (3C), 43.1 (3C), 36.5 (3C), 30.5 (3C), 25.4 (3C), 19.4 (3C). HRMS (ESI) m/z:
[M+H]
+
calcd for C
69
H
88
N
3
O
15
1198.6210; found 1198.6198.
Aza-cryptophane anti-3 and syn-4. To a mixture of acetic acid (600 mL) and perchloric
acid (200 mL) under an argon atmosphere, was added dropwise (12 hours) a solution of the
cryptophane precursor 11 (8.0 g, 6.67 mmol) in acetic acid (200 mL). After the end of the
addition, the mixture was further stirred for 6 hours and poured into water (1 L). The product
was extracted with CH
2
Cl
2
(2 x 1 L) and the organic layer was washed with water then with a
saturated solution of Na
2
CO
3
to remove traces of perchloric acid. The organic layer was dried
over sodium sulfate and the solvent was evaporated under vacuum. Two cryptophane
derivatives were identified by
1
H NMR spectroscopy from this crude product. These two
compounds were separated by column chromatography on silica gel.
The first eluted product was identified as the compound anti-3. In order to isolate this
diastereomer, the reaction described above was repeated three times. The crude product was
purified on silica gel (CH
2
Cl
2
88% - acetone 12%). The different fractions were evaporated by
rotatory evaporation to give a product, which was precipitated in Et
2
O. Recrystallisation in
CH
2
Cl
2
/ethanol to give rise to anti-3 (205 mg, 1.2%) as a beige solid. mp 220°C (decomp);
1
H
NMR (400 MHz, CDCl
3
, 25 °C): δ 6.72 (s, 3H), 6.67 (s, 3H), 6.54 (s, 3H), 6.35 (s, 3H), 4.60
(d, 3H, J = 13.5 Hz), 4.57 (d, 3H, J = 13.5 Hz,), 4.26-4.21 (m, 1H), 4.00-3.95 (m, 3H), 3.79 (s,
9H), 3.77 (s, 9H), 3.38-3.33 (m, 6H), 3.37 (d, 3H, J = 13.9 Hz), 3.30 (d, 3H, J = 13.8 Hz).
13
C{
1
H} NMR (101 MHz, CDCl
3
, 25 °C): δ 149.2 (3C), 146.4 (3C), 146.3 (3C), 136.2 (3C),
133.5 (3C), 132.3 (3C), 131.4 (3C), 128.7 (3C), 121.7 (3C), 114.2 (3C), 113.0 (3C), 111.4 (3C),
70.1 (3C), 55.7 (3C), 55.3 (3C), 44.4 (3C), 36.3 (3C), 36.0 (3C). UV (THF) log
310 nm
(3.7, shoulder), 287 nm (4.15), 229 nm (4.85). HRMS (ESI) m/z: [M+H]
+
calcd for C
54
H
58
N
3
O
9
28
892.4168; found 892.4164. Rf (DCM - acetone (85:15)): 0.55. (-)-MP-anti-3’, []
D
25
= - 230.7
(c = 0.26, CH
2
Cl
2
); (+)-PM-anti-3’, []
D
25
= + 222.3 (c = 0.21, CH
2
Cl
2
).
The second eluted product was identified as the syn-4 compound. It was purified on
silica gel (CH
2
Cl
2
85%- acetone 15%). It was then precipitated in Et
2
O and recrystallized in
dichloromethane/ethanol to give rise to a beige solid (330 mg, 6%), which was identified as the
syn-4 derivative. mp 180°C (decomp);
1
H NMR (400 MHz, CDCl
3
, 25 °C): δ 6.75 (s, 3H), 6.62
(s, 3H), 6.52 (s, 3H), 6.36 (s, 3H), 4.60 (d, 6H, J = 13.8 Hz), 4.15 (d, 3H, J = 9,6 Hz,), 3.79 (s,
9H), 3.74 (s, 9H), 3.58-3.36 (m, 9H), 3.40 (d, 3H, J = 13.8 Hz) 3.30 (d, 3H, J = 13.8 Hz).
13
C{
1
H} NMR (101 MHz, CDCl
3
, 25 °C): δ 149.8 (3C), 147.7 (3C), 146.7 (3C), 136.5 (3C),
134.0 (3C), 132.3 (3C), 131.4 (3C), 128.3 (3C), 120.8 (3C), 112.9 (3C), 112.2 (3C), 112.1 (3C),
71.2 (3C), 55.9 (3C), 55.7 (3C), 43.5 (3C), 36.6 (3C), 36.4 (3C). UV (THF) log
315 nm
(4.04; shoulder), 296 nm (4.2), 255 nm (4.45, shoulder). HRMS (ESI) m/z: [M+H]
+
calcd for
C
54
H
58
N
3
O
9
892.4168; found 892.4188. Rf (DCM - acetone (85:15)): 0.4. (-)-PP-1, []
D
25
= -
42.0 (c = 0.29, CH
2
Cl
2
); (+)-MM-1, []
D
25
= + 45.2 (c = 0.30, CH
2
Cl
2
).
Supporting Information.
1
H and
13
C spectra of compounds 3-4, 7-8 and 10-11. Analytical
HPLC of compound anti-3 and syn-4.
1
H NMR spectra of compounds (-)-MP-3, (+)-PM-3, (-)-
PP- 4 and (+)-MM-4. Optical rotations of compounds (-)-MP-3, (+)-PM-3, (-)-PP- 4 and (+)-
MM-4. IR and VCD spectra of (+)-PP-1 and (+)-PM-3. ECD spectra of compounds (-)-MP-3,
(+)-PM-3, (-)-PP- 4 and (+)-MM-4. One-scan
129
Xe NMR spectra of compounds anti-3 and
syn-4. Geometries of anti-3 and syn-4 conformers used for IR/VCD spectra calculations.
AUTHOR INFORMATION
29
Corresponding author
*E-mail: [email protected]; [email protected]
ORCID
Thierry Brotin: 0000-0001-9746-4706
Nicolas De Rycke: 0000-0001-6487-1030
Patrick Berthault: 0000-0003-4008-2912
Thierry Buffeteau: 0000-0001-7848-0794
Nicolas Vanthuyne: 0000-0001-7848-0794
Martin Doll: 0000-0001-9254-1821
Notes
The authors declare no competing financial interest.
ACKNOWLEDGEMENTS
The French National Research Agency (ANR) is acknowledged for financial support (Project
ANR19-CE19-0024 PHOENIX).
(1) a) Cram, D. J.; Cram, J. M. Host-Guest Chemistry. Complexes Between Organic
Compounds Simulate the Substrate Selectivity of Enzymes. Science, 1974, 183, 803-809.
b) Cram, D. J.; Cram, J. M. Design of Complexes Between Synthetic Hosts and Organic
30
Guests. Account of Chem. Res. 1978, 8-14. c) Chao, Y.; Weisman. G. R.; Sogah, G. D.
Y.; Cram. D. J. Host Guest Complexation. 21. Catalysis and Chiral Recognition through
Designed Complexation of Transition States in Transacylations of Amino Ester Salts. J.
Am. Chem. Soc. 1979, 101, 4948-4958. d) Cram. D. J. The Design of Molecular Host,
Guest, and their Complexes. Science, 1988, 240, 760-767.
(2) a) Collet, A. Cyclotriveratrylene and Cryptophane. Tetrahedron, 1987, 4, 24, 5725-5759.
b) Brotin, T.; Dutasta, J.-P. Cryptophanes and their Complexes: Present and Future.
Chem. Rev. 2009, 109, 88-130.
(3) a) Garel, L.; Dutasta, J.-P.; Collet, A. Complexation of Methane and Chlorofluorocarbons
by Cryptophane-A in Organic Solution. Angew. Chem. Int. Ed. 1993, 32, 1169-1171. b)
Bartik, K.; Luhmer, M.; Dutasta, J.-P.; Collet, A.; Reisse, J.
129
Xe and
1
H NMR Study of
the Reversible Trapping of Xenon by Cryptophane-A in Organic Solution. J. Am. Chem.
Soc. 1998, 120, 784-791. c) Brotin, T.; Montserret, R.; Bouchet, A.; Cavagnat, D.;
Linares, M.; Buffeteau, T. High Affinity of Water-Soluble Cryptophanes for Cesium
Cation. J. Org. Chem. 2012, 77, 1198-1201.
(4) See for instance: a) Baydoun, O.; De Rycke, N.; Léonce, E.; Boutin, C.; Berthault, P.;
Jeanneau, E.; Brotin, T. Synthesis of Cryptophane-223 Type Derivatives with Dual
Functionalization. J. Org. Chem. 2019, 84, 9127-9137. b) Dubost, E.; Kotera, N.; Garcia-
Argote, S.; Boulard, Y.; Léonce, E.; Boutin, C.; Berthault, P.; Dugave, C.; Rousseau, B.
Synthesis of a Functionalizable Water-Soluble Cryptophane-111. Org. Lett. 2013, 15,
2866-2868. c) Taratula, O.; Hill, P. A.; Khan, N. S.; Dmochowski, I. J. Shorter Synthesis
of Trifunctionalized Cryptophane-A Derivatives. Org. Lett. 2011, 13, 1414-1417. d)
Wong, T-H.; Chang, J-C.; Lai, C-C.; Liu, Y-H. ; Peng, S-M. ; Chiu, S-H.
Hemicarceplexes Modify the Solubility and Reduction Potentials of C
60
. J. Org. Chem.
2014, 79, 3581-3586.
31
(5) a) Little, M. A.; Donkin, J.; Fisher, J.; Halcrow, M. A.; Loder, J.; Hardie, M. J. Synthesis
and Methane-Binding Properties of Disulfide-Linked Cryptophane-0.0.0. Angew. Chem.
Int. Ed. 2012, 51, 764-766. b) Brégier, F.; Lavalle, J.; Chambron, J.-C. Capping α-
Cyclodextrin with Cyclotriveratrylene by Triple Disulfide-Bridge Formation. Eur. J. Org.
Chem. 2013, 13, 2666-2671.
(6) Bohle, D. S.; Stasko, D. J. Salicylaldiminato Derivatives of Cyclotriveratrylene: Flexible
Strategy for New Rim-Metalated CTV Complexes. Inorg. Chem. 2000, 39, 5768-5770.
(7) Gabard, J.; Collet, A. Synthesis of a (D
3
)-Bis(cyclotriveratrylenyl) Macrocage by
Stereospecific Replication of a (C
3
)-Subunit. J. Chem. Soc. Chem. Comm. 1981, 1137-
1139.
(8) Brotin, T.; Jeanneau, E.; Berthault, P.; Léonce, E.; Pitrat, D.; Mulatier, J.-C. Synthesis of
Cryptophane-B: Crystal Structure and study of its complex with Xenon. J. Org. Chem.
2018, 83, 14465-14471.
(9) a) IUPAC Tentative Rules for the Nomenclature of Organic Chemistry. Section E.
Fundamental Stereochemistry. J. Org. Chem. 1970, 35, 2849-2867. b) Collet, A.; Gabard,
G.; Jacques, J.; Césario, M.; Guilhem, J.; Pascard, C. Synthesis and Absolute
Configuration of Chiral (C
3
) Cyclotriveratrylene Derivatives. Crystal Structure of (M)-(-
)2,7,12-Triethoxy-3,8,13-tris-[(R)-1-methoxycarbonylethoxy]-10,15-dihydro-5H-
tribenzo[a,d,g]-cyclononene. J. Chem. Soc. Perkin Trans. 1 1981, 1630-1638.
(10) Garcia, C.; Malthête, J.; Collet, A. Key Intermediates in Cyclotriveratrylene Chemistry.
Synthesis of New C3-cyclotriveratrylenes with Nitrogen Substituents. Bull. Soc. Chim.
Fr., 1993, 130, 93-95.
(11) Brotin, T.; Vanthuyne, N.; Cavagnat, D.; Ducasse, L., Buffeteau, T. Chiroptical
Properties of Nona- and Dodecamethoxy Cryptophanes. J. Org. Chem. 2014, 79, 6028-
6036.
32
(12) We use the following convention to describe the aza-compounds: The first stereo-
descriptor is related to the oxygen-containing CTB unit and the second stereo-descriptor
is related to the nitrogen-containing CTB unit.
(13) See for instance: a) Brotin, T.; Roy, V.; Dutasta, J.-P. Improved Synthesis of Functionals
CTVs and Cryptophanes Using Sc(OTf)
3
as catalyst. J. Org. Chem. 2005, 70, 6187 - 6195.
b) Long, A.; Jean. M.; Albalat, M.; Vanthuyne, N.; Giorgi, M.; Gorecki, M.; Dutasta, J.-
P.; Martinez, A. Synthesis, Resolution, and Chiroptical Properties of Hemicryptophane
cage Controlling the Chirality of Propeller Arrangement of a C
3
Triamide Unit. Chirality,
2019, 31, 910-916. c) Long, A.; Colomban, C.; Jean. M.; Albalat, M.; Vanthuyne, N.;
Giorgi, M.; Di Bari, L.; Gorecki, M.; Dutasta, J.-P.; Martinez, A. Enantiopure C
1
-
Cyclotriveratrylene with a Reversed Spatial Arrangement of the Substituents. Org. Lett.
2019, 21, 160-165.
(14) Daugey, N.; Brotin, T.; Vanthuyne, N.; Cavagnat, D.; Buffeteau, T. Raman Optical
Activity of Enantiopure Cryptophane. J. Phys. Chem. B 2014, 118, 5211-5217.
(15) Canceill, J.; Collet, A.; Gottarelli, G.; Palmieri, P. Synthesis and Exciton Optical Activity
of D
3
-Cryptophanes. J. Am. Chem. Soc. 1987, 109, 6454-6464.
(16) Taratula, O.; Hill, P. A.; Carroll, P. J.; Dmochowski, I. J. Crystallographic Observation
of ‘Induced Fit’ in a Cryptophane Host-Guest Model System. Nat. Commun. 2010, 148,
1-7.
(17)
Fogarty, H. A.; Berthault, P.; Brotin, T.; Huber, G.; Desvaux, H.; Dutasta, J.-P. A
Cryptophane Core Optimized for Xenon Encapsulation. J. Am. Chem. Soc.,
2007
,
129, 10332-10333.
(18) Spence, M. M.; Rubin, S. M.; Dimitrov, I. E.; Ruiz, J.; Wemmer, D.; Pines, A.; Yao, S.
Q.; Tian, F.; Schultz, P. G. Functionalized Xenon as a Biosensor. Proc. Natl. Acad. U.S.A,
2001, 98, 10654-10657.
33
(19) Brotin, T.; Martinez, A.; Dutasta, J.-P. Water-Soluble Cryptophanes: Design and
Properties. In Calixarenes and Beyond; Neri, P., Sessler, J. L., Wang, M.-X., Eds.;
Springer International Publishing: Cham, 2016; pp 525-557.
(20) a) Jayapaul, J.; Schröder; L. Molecular Sensing with Host Systems for Hyperpolarized
129Xe. Molecules, 2020, 25, 4627-4723. b) Mari, E.; Berthault, P.
129
Xe NMR Based
Sensor: Biological Applications and Recent Methods. Analyst, 2017, 142, 3298-3308. c)
Klippel, S.; Döpfert, J.; Jayapaul, J.; Kunth, M.; Rossella, F.; Schnurr, M.; Witt, C.;
Freund, C.; Schröder, L. Cell Tracking with Caged Xenon: Using Cryptophanes as MRI
Reporters Upon Cellular Internalization. Angew. Chem. 2014, 126, 503-506. d) Schröder,
L. Xenon for NMR Biosensing – Inert but Alert. Physica Medica, 2013, 29, 3-16. e)
Taratula, O.; Dmochowski, I. J. Functionalized
129
Xe Contrast Agents for Magnetic
Resonance Imaging. Current Opinion in Chemical Biology 2010, 14, 97-104. f) Berthault,
P.; Huber, G.; Desvaux, H. Biosensing Using Laser-Polarized Xenon NMR/MRI. Progr.
NMR Spectrosc. 2009, 55, 35-60. g) Hilty, C.; Lowery, T. L. Wemmer, D. E.; Pines, A.
Spectrally Resolved Magnetic Resonance Imaging of a Xenon Biosensor. Angew. Chem.
Int. Ed. 2006, 45, 70-73.
(21) Schroder, L.; Lowery, T. J.; Hilty, C.; Wemmer, D.; Pines, A. Molecular Imaging Using
a Targeted Magnetic Resonance Hyperpolarized Biosensor. Science, 2006, 314, 446-449.
(22) Dubost, E.; Dognon, J.-P.; Rousseau, B.; Milanole, G.; Dugave, C.; Boulard, Y.; Léonce,
E.; Boutin, C.; Berthault, P. Understanding a Host–Guest Model System through
129
Xe
NMR Spectroscopic Experiments and Theoretical Studies. Angew. Chem. Int. Ed. 2014,
53, 9837-9840.
(23) Chauvin, C. ; Liagre, L. ; Boutin, C. ; Mari, E. ; Léonce, E. ; Carret, G. ; Coltrinari, B. ;
Berthault. P. Spin-Exchange Optical Pumping in a Van”. Review of Scientific
Instruments 87 (2016) 016105.
34
(24) a) Buffeteau, T.; Lagugné-Labarthet, F.; Sourrisseau, C. Vibrational Circular Dichroism
in General Anisotropic Thin Solid Films: Measurement and Theoretical Approach. Appl.
Spectrosc. 2005, 59, 732-745. b) Brotin, T.; Daugey, N.; Vanthuyne, N.; Jeanneau, E.;
Ducasse, L., Buffeteau, T. Chiroptical Properties of Cryptophane-223 and -233
Investigated by ECD, VCD, and ROA Spectroscopy. J. Phys. Chem. B 2015, 119, 8631-
8639.
(25) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman,
J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A. et al. (see Supporting
Information for full list of authors) Gaussian 09, revision A.1, Gaussian Inc., Wallingford
CT, 2009.
