Xe@cryptophane complexes with C2 symmetry: Synthesis and investigations by 129Xe NMR of the consequences of the size of the host cavity for xenon encapsulation

Article abstract of DOI:10.1002/ejoc.200390153

New cryptophanes-223 (1), -233 (2), and -224 (3) with C₂ symmetry, bearing different linkers connecting the two cyclotriveratrylene units were synthesized following a multistep procedure. The formation of the xenon@cryptophane complexes was investigated by ¹²⁹Xe NMR spectroscopy. Cryptophanes-223 (1) and -233 (2) complex xenon efficiently in 1,1,2,2-[D₂]tetrachloroethane solution with somewhat lower binding constants (K = 2810 M^-1 and 810 M^-1 respectively at 278 K) than that observed with cryptophane-A (K = 3900 M^-1). The free and bound guests were observed at room temperature under slow-exchange conditions on the ¹²⁹Xe NMR timescale. A linear relationship was observed between the binding constants and the internal volume of the hosts. Under the same conditions, the Xe@cryptophane-224 [Xe3] complex underwent a fast-exchange process on the NMR timescale. The decomplexation activation energies E_a, and the associated parameters ΔH? and ΔS?, were determined for the Xe1 and Xe2 complexes from variable temperature 1D-EXSY experiments. These investigations emphasized the importance of the size of the internal cavity of the cryptophanes for guest encapsulation, and the extreme sensitivity of the ¹²⁹Xe chemical shift toward slight structural modifications of the atom's environment. ( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim, Germany, 2003).

Full text of DOI:10.1002/ejoc.200390153

FULL PAPER
Xe@cryptophane Complexes with C2 Symmetry: Synthesis and Investigations
by 129Xe NMR of the Consequences of the Size of the Host Cavity for Xenon
Encapsulation
Thierry Brotin[a] and Jean-Pierre Dutasta*[a]
Keywords: Supramolecular chemistry / Cryptophanes / Host-guest systems / NMR spectroscopy / Xenon
New cryptophanes-223 (1), -233 (2), and -224 (3) with C2
symmetry, bearing different linkers connecting the two cyc-
lotriveratrylene units were synthesized following a multi-
step procedure. The formation of the xenon@cryptophane
complexes was investigated by 129Xe NMR spectroscopy.
Cryptophanes-223 (1) and -233 (2) complex xenon efficiently
in 1,1,2,2-[D2]tetrachloroethane solution with somewhat lo-
der the same conditions, the Xe@cryptophane-224 [Xe3]
complex underwent a fast-exchange process on the NMR
timescale. The decomplexation activation energies Ea, and
the associated parameters ∆H‡ and ∆S‡, were determined for
the Xe1 and Xe2 complexes from variable temperature 1D-
EXSY experiments. These investigations emphasized the im-
portance of the size of the internal cavity of the cryptophanes
wer binding constants (K = 2810
at 278 K) than that observed with cryptophane-A (K = 3900
M M
−1 and 810 −1 respectively for guest encapsulation, and the extreme sensitivity of the
129Xe chemical shift toward slight structural modifications of
M
−1). The free and bound guests were observed at room tem- the atom’s environment.
perature under slow-exchange conditions on the 129Xe NMR
timescale. A linear relationship was observed between the
binding constants and the internal volume of the hosts. Un-
( Wiley-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim,
Germany, 2003)
Introduction
(K ഠ 5Ϫ10 Ϫ1 at 278 K in 1,1,2,2-tetrachloroethane).[7]
The 129Xe NMR spectrum of a mixture of xenon and crypto-
phane-E revealed only one broad signal located at high
frequency, corresponding to free xenon in solution under
fast-exchange conditions on the NMR timescale. This result
shows that a slight change in the volume of the molecular
cavity of the cryptophanes has a dramatic effect on the
complexing properties towards xenon. Such a dependence
of guest recognition on the internal volume of hosts has
already been reported for other supramolecular systems
and plays an important role in the encapsulation of neutral
molecules inside molecular cavities.[8]
In this paper, we report the synthesis of the new crypt-
ophanes-223 (1), -233 (2) and -224 (3) bearing different lin-
kers, where -223, -233 and -224 indicate the number of
methylene groups in each of the three linkers [e.g. -223
stands for two (CH2)2 and one (CH2)3 linkers]. Accordingly,
cryptophane-A and -E should be labeled -222 and -333 re-
spectively. However, we will keep the former nomenclature
throughout this paper. We found that cryptophane-223
and -233, for which the sizes of the internal cavities lie be-
tween those of cryptophane-A and cryptophane-E, complex
Small neutral molecules are known to interact weakly
with their environment, mainly through van der Waals
forces. Therefore, the design of original organic receptors
able to encapsulate guests such as helium or xenon atoms
is still a challenging topic.[1Ϫ4] Cryptophane-A, which be-
longs to a well-known family of supramolecular hosts, has
demonstrated its ability to encapsulate monatomic xenon
efficiently with a surprisingly high binding constant, K ഠ
3900 Ϫ1 at 278 K.[5] The extraordinary sensitivity of the
129Xe chemical shift to the atom’s environment, and par-
ticularly to the confined space of the cryptophane cavity,
allowed us to discriminate between cryptophane-A and its
deuterated congeners.[6] This exceptional property suggests
that the size of the internal cavity of cryptophane-A is well
suited to encapsulate xenon, thus maximizing hostϪguest
interactions. This was further supported by studying xenon
complexation with cryptophane-E, which possesses three
propylene bridges connecting the two cyclotriveratrylene
units, instead of the ethylene bridges in cryptophane-A.
Consequently, cryptophane-E has a larger inner cavity than
cryptophane-A, and complexes monoatomic xenon poorly xenon efficiently in solution under slow-exchange condi-
tions on the NMR timescale. We also determined the de-
[a]
´
complexation activation energy and the associated para-
meters ∆H‡ and ∆S‡ of the new Xe@cryptophane com-
plexes from a series of 129Xe 1D-EXSY NMR experiments
performed at several temperatures.
´ ´
´
Stereochimie et Interactions Moleculaires, Ecole Normale
´
Superieure de Lyon, UMR CNRS no. 5532,
´
46 Allee d’Italie, 69364 Lyon 07, France
Fax: (internat.) ϩ 33-4/72728483
E-mail: dutasta@ens-lyon.fr
Eur. J. Org. Chem. 2003, 973Ϫ984  2003 WILEY-VCH Verlag GmbH & Co. KGaA, 69451 Weinheim 1434Ϫ193X/03/0306Ϫ0973 $ 20.00ϩ.50/0
973

T. Brotin, J.-P. Dutasta
FULL PAPER
their precursors during LSIMS experiments, they can be
prepared using standard synthetic procedures.[10]
Vanillyl alcohol (4) was allowed to react with allyl bro-
mide in acetone in the presence of potassium carbonate to
give the protected compound 5 (Scheme 1). This is the key
compound for generating cyclotriveratrylene 6 in two steps,
in fairly good yields, according to a known procedure.[11]
Vanillyl alcohol was also used to prepare compounds 7Ϫ9
bearing bromoalkyl chains with different lengths. Following
the standard procedure using potassium carbonate in ace-
tone and the appropriate dibromoalkane, they were ob-
tained from 4 in low (25%, 7)[12,13] to good yields (65%,
8;[12,13] 75%, 9). The tetrahydropyranyl ether (OTHP) func-
tions, introduced as protecting groups in the cryptophane
precursors, considerably facilitated their purification by col-
umn chromatography on silica gel and did not alter the fi-
nal cyclization step. Thus, reaction of 7 and 8 with dihy-
dropyran and pyridinium p-toluenesulfonate (PPTS) in a
mixture of THF and dichloromethane led to compounds 10
and 11, respectively in 80% yield. The corresponding iodin-
ated derivatives were obtained in quantitative yield by treat-
ment with an excess of sodium iodide in acetone to give
compounds 12؊16 (Scheme 1).[9,12,13]
The preparation of cryptophanes 1؊3 required a slightly
different strategy from that used for the cryptophanes with
D3 symmetry. In the present case we needed to introduce
different substituents on the cyclotriveratrylene 6. This
could be achieved only through a stepwise synthesis intro-
ducing the different substituents successively in order to ob-
tain the desired cryptophanes (Scheme 2). Such a strategy
was successfully applied to introduce a single deuterated
methyl group or two deuterated alkyl chains in a crypto-
phane structure.[9] Reaction of 6 with two equivalents of
cesium carbonate in DMF and the appropriate iodinated
compounds 12, 13 and 14 afforded compounds 17 (30%),[9]
18 (32%) and 19 (33%) respectively. The terminal benzyl
alcohol function was kept unprotected in order to facilitate
Results and Discussion
Synthesis and Characterization
The synthesis of the new cryptophanes 1؊3 was achieved
by following the experimental procedure previously de-
scribed for the synthesis of deuterated cryptophane-A.[9]
This method requires the generation of the two cyclotrivera-
trylene caps at different stages of the synthesis, and usually
gives rise to cryptophanes in moderate to good yields, de-
pending on their chemical structure. The formation of the
cryptophanes from their precursors was predicted from
mass spectrometry experiments. Indeed, we have recently
demonstrated that when cryptophanes are formed from
Scheme 1
974
Eur. J. Org. Chem. 2003, 973Ϫ984

Xe@cryptophane Complexes with C2 Symmetry
FULL PAPER
Scheme 2
the chromatographic separation of the desired material on
silica gel. Side products were mainly the di- and trisubsti-
tuted derivatives, which were obtained in similar yields.
Cryptophane precursors 20Ϫ22 were prepared from 17؊19
by reaction with compounds 15 or 16 in good yields (72 to
79%) after column chromatography on silica gel.
The new cryptophanes 1, 2 and 3 were then obtained
from their precursors 21, 20 and 22 respectively, by treat-
ment in formic acid. The yields were 35Ϫ40% after purifica-
tion, significantly lower than that obtained with crypto-
phane-A (70%) prepared under the same experimental
conditions. Presumably, the dissymmetry of the precursors
did not favor the cyclization step and therefore increased
the formation of side-products.
The 1H NMR spectra (Figure 1) recorded in chloroform,
and the 13C NMR spectra of compounds 1؊3 are in good
1
agreement with a structure with C2 symmetry. In the H
NMR spectra of 1؊3, the aromatic and methyl protons give
six and three signals respectively. In addition, for each com-
pound different patterns were observed for the axial and
the equatorial protons of the methylene bridges, as a con-
sequence of the dissymmetry of the molecules.
Nevertheless, these observations are not sufficient to fully
describe the structure of compounds 1؊3. Indeed, the strat-
egy used to synthesize the cryptophanes can lead to the syn
Figure 1. 1H NMR spectra of cryptophanes 1Ϫ3 in deuterated
(achiral, Cs symmetry) and/or the anti isomers (chiral, C2 chloroform at 293 K (* impurity from the solvent)
Eur. J. Org. Chem. 2003, 973Ϫ984
975

T. Brotin, J.-P. Dutasta
FULL PAPER
1
symmetry), undistinguishable from their H NMR spectra. 177.0 ppm, respectively (vs. δ ϭ 156.0 ppm for crypt-
The stereochemistry of compounds 1 and 2 was determined ophane-A). At 238 K the peaks are narrower and the chem-
by chiral HPLC using a Chiralpak OTϩ column. This tech- ical shift differences between free and bound xenon in 1 and
nique was successfully used to discriminate the two enanti- 2 are δ ϭ 201.0 and δ ϭ 211.0 ppm, respectively. The shield-
omers of the anti isomer of cryptophane-A and related ing environment of the benzene rings of the host molecule
compounds.[14,15] Both compounds 1 and 2 exhibite two is not the major contributor to the large chemical shift dif-
close peaks of equal intensity indicating that we isolated the ference observed between the free and bound xenon. As
anti isomer (C2 symmetry). HPLC analysis of 3 showed reported by Bartik et al., intermolecular interactions, and
only one broad signal, presumably due to a very low resolu- particularly van der Waals interactions, are by far the most
tion, which prevented the characterization of both enanti- important effect because of the large polarizable electron
omers.[16] However, the syn isomers of cryptophanes bear- cloud of the xenon atom.[5] However, because the environ-
ing O(CH2)2O linkers have never yet been observed, and we ment of the bound guest within the three cryptophanes is
tentatively assigned the anti configuration to cryptophane
3. Only an X-ray structural determination could ascertain
the stereochemistry of host 3. Unfortunately, so far we have
not obtained crystals suitable for X-ray analysis.
very similar, an additional factor has to be considered to
explain the different chemical shifts observed for each com-
plex.
It is well-known that the xenon chemical shift is highly
sensitive to slight variations of the volume of the molecular
cavity, which defines the close surroundings of the xenon
atom.[17] The differences observed for the resonance of the
xenon encapsulated within the cavity of each cryptophane
host probably arise from a change of the cavity size. Bound
xenon should be less confined in 2 than in cryptophane-A,
explaining the low frequency shift observed with Xe2. It is
obvious that a change in the length of the O(CH2)nO linkers
will modify the size of the windows of the cryptophanes
and consequently should affect the dynamics of the com-
plexation process. This point will be discussed later. The
129Xe NMR spectrum of a solution containing xenon and
cryptophane-224 (3) was recorded at 233 K (Figure 2, c).
The spectrum showed only one broad signal at δ ϭ
245 ppm, the value for free xenon. The absence of a reso-
129Xe NMR Studies of the Xe@cryptophane Complexes
The 129Xe NMR spectra of the Xe@cryptophane-223,
-233 and -224 complexes were first studied separately at sev-
eral temperatures in 1,1,2,2-[D2]tetrachloroethane (Fig-
ure 2). At room temperature, we observed the resonance of
the free and bound xenon in molecules 1 and 2 (slow-ex-
change conditions). Encapsulation of xenon by crypt-
ophane-233 (2) seems surprising since the structure of this
molecule is probably closer to the structure of cryptophane-
E than that of cryptophane-A. We also noticed a large
chemical-shift difference for the resonance of bound xenon
encapsulated within the three cryptophanes. The difference
becomes even larger upon decreasing the temperature. At
293 K the chemical shift differences between free and nance for bound xenon suggested that we were under fast-
bound xenon in molecules 1 and 2 are δ ϭ 164.0 and δ ϭ exchange conditions on the NMR timescale and the
Figure 2. 129Xenon NMR spectrum of the of Xe@cryptophane-233 complexes [Xe2] in 1,1,2,2-[D2]tetrachloroethane; the 129Xe NMR
signals of xenon encapsulated into cryptophane-A and cryptophane-223 (1) are also shown for comparison; a) 293 K, b) 238 K; c) 129Xe
NMR spectrum of a solution containing cryptophane-224 (3) in 1,1,2,2-[D2]tetrachloroethane solution at 233 K (only the resonance of
the unbound xenon is observed under fast-exchange conditions on the NMR timescale)
976
Eur. J. Org. Chem. 2003, 973Ϫ984

Xe@cryptophane Complexes with C2 Symmetry
FULL PAPER
3
˚
Table 1. Volume V ( A ) of the internal cavities of cryptophanes-A, -E and 1؊3
Cryptophane-A
95.0
Cryptophane-223 (1)
Cryptophane-233 (2)
Cryptophane-224 (3)
109.9
Cryptophane-E
120.9
3
˚
V ( A )
102.1
117.2
broadening observed for the free xenon line was due to
As expected, the two samples showed the presence of
xenon going into 3 with a rather short residence time. To three signals, for the free xenon and the two corresponding
confirm these observations the sizes of the internal cavities Xe@cryptophane complexes. In the first sample, the inten-
of the new cryptophanes 1Ϫ3 were calculated by molecular sity of the signal for Xe1 is significantly lower than that for
modeling.[18] The results reported in Table 1, represent the Xe@cryptophane-A, suggesting a lower binding constant
volume of the internal cavity of the minimized structure of for 1 than for cryptophane-A. Similarly, in the second
the cryptophanes in the gas phase. As expected, the dimen- sample, a significant decrease of the signal of Xe2 is ob-
sions of the internal cavity increase on going from crypt- served with respect to Xe1. This is in good agreement with
ophane-A to cryptophane-E except for cryptophane-224 an increase of the volume of the internal cavity of hosts 1
(3), where they are between those of cryptophane-223 (1) and 2 with respect to cryptophane-A. The association con-
and cryptophane-233 (2).
stants for the new complexes Xe1 (K ϭ 2810 Ϫ1) and Xe2
This result is surprising since we have observed that in (K ϭ 810 Ϫ1) were obtained from the peak intensities of
solution xenon binding to cryptophane-224 (3) is not as the corresponding xenon signals and the known binding
efficient as to cryptophane-233 (2). The conclusion is prob- constant for Xe@cryptophane-A (K ϭ 3900 Ϫ1).
ably that the O(CH2)4O linker exists in several conforma-
These results show that a slight modification in the struc-
tions close in energy, which can significantly change the size ture of the hosts (volume of the internal cavity) significantly
of the internal cavity. The increased flexibility due to the affects their binding properties. Additionally, we observed
O(CH2)4O linker prevents the formation of a stable Xe3 a linear relationship between the measured binding con-
complex. Additionally, with the larger cryptophane 3, the stants and the volume of the molecular cavity (Figure 4).
solvent may compete with the guest. However, this is diffi- This interesting feature suggests the possibility of predicting
cult to verify experimentally owing to the lack of available the complexing behavior toward xenon of cryptophanes (or,
noncompeting solvents.
probably, any other molecules accommodating xenon). This
Binding constants of the new Xe1 and Xe2 complexes may have important consequences, especially for designing
were measured by competition experiments. A solution con- new efficient hosts for practical applications such as biosen-
taining equimolar amounts of cryptophane-A and 1 was sors, a concept that has recently been described using hyper-
prepared in 1,1,2,2-tetrachloroethane in the presence of an polarized xenon.[19]
excess of xenon. The 129Xe NMR spectrum was then re-
corded at 278 K (Figure 3, a). Similarly, a second sample
Variable-Temperature 129Xe 1D-EXSY NMR Experiments
was prepared by mixing equimolar amounts of crypto-
The slow-exchange conditions observed for the com-
phanes 1 and 2 (Figure 3, b).
plexes Xe1 and Xe2 allowed us to study the exchange dy-
Figure 3. 129Xenon NMR spectra recorded at 278 K in 1,1,2,2-tetrachloroethane of a solution containing an equimolar amount of a)
cryptophane-A and cryptophane-223 (1); b) cryptophane-223 (1) and cryptophane-233 (2)
Eur. J. Org. Chem. 2003, 973Ϫ984
977

T. Brotin, J.-P. Dutasta
(1)
FULL PAPER
The matrix L is a function of the rate constants kij and
the longitudinal relaxation time Tij of each site for a given
kinetic model. We have previously clearly demonstrated
from 2D-EXSY and 1D-EXSY NMR experiments that
xenon exchange between cryptophanes takes place exclus-
ively by xenon atoms travelling through solution and that
the second process involving xenon exchange by direct colli-
sion between the hosts can be neglected under our experi-
mental conditions. Thus, we propose the following pseudo-
first-order kinetic model (Scheme 3).
Figure 4. Plot of the measured binding constant K (Ϫ1) of crypto-
phanes-A, -E, 1, and 2 as a function of the volume of the internal
cavity (r2 ϭ 0.996)
namics in solution between the bound and free xenon
sites.[20] The decomplexation activation energy Ea and the
associated activation parameters ∆H‡ and ∆S‡ can be calcu-
lated from the rate constants determined at several temper-
atures. We successfully applied this to cryptophane-A and
one of its deuterated congeners by using 129Xe 1D-EXSY
NMR spectroscopy.[9] We preferred the 129Xe 1D-EXSY
technique to 2D-EXSY for several reasons. In particular,
the low number of exchanging sites, the better precision in
the determination of the rate constants and the substantial
decrease in the instrument time compared to 2D-EXSY,
make 1D-EXSY experiments very useful for measuring the
decomplexation activation energy.
We used two different methods to measure the activation
parameters of the new complexes. In a first approach, we
determined the activation parameters from a solution con-
taining equimolar amounts of cryptophane-A and 1. This
method has the advantage of estimating the activation para-
meters of the new cryptophane with cryptophane-A as an
internal standard, since its activation parameters are
known. In addition it gives a good indication of the preci-
sion of our results. On the other hand, it complicates the
fitting procedure since more variables are needed to fit the
Scheme 3. Pseudo-first-order kinetic model used to describe the
1D-EXSY experiments for a sample containing an equimolar
amount of cryptophane-A and cryptophane-223 (1) in the presence
of an excess of xenon.
With this model we can describe the exchange dynamics
of xenon between cryptophanes as two independent systems
involving free xenon and cryptophane-A or free xenon and
host 1, respectively. Moreover, we have also clearly shown
that the longitudinal relaxation time T1, which is the same
for all sites, is several orders of magnitude greater than the
rate constants and can be neglected in the matrix L. The L
matrix can therefore be simplified to the form given in
Equation (2).[21]
(2)
The equilibrium equation also implies k1,3/k3,1 ϭ M01/
experimental data. We could not use a sample containing M03 and k3,2/k2,3 ϭ M02/M03 where M01, M02 and M03 are re-
equimolar amounts of 1 and 2 to obtain accurate rate con- spectively the initial magnetization (τm ϭ 0) of xenon
stants for 2. The large difference of the intensities observed bound to cryptophane-A, xenon bound to cryptophane-223
between the two bound states would have led to inaccurate and free xenon in solution. For better precision we meas-
data because of the low signal-to-noise ratio. Thus, in a se- ured M01, M02 and M03 by integrating the corresponding sig-
cond approach, we studied cryptophanes 1 and 2 separately. nals in the 1D 129Xe NMR spectrum at a given temperature.
The first method was thus only applied to molecule 1 We obtained the whole set of rate constants by irradiating
(Experiment 1). A sample containing an equimolar amount independently the resonance corresponding to the free
of 1 and cryptophane-A in the presence of an excess of xenon in solution and the two resonances corresponding to
xenon was prepared and sealed under vacuum. The magnet- the encapsulated xenon in both cryptophanes. This does not
ization of the three sites (free xenon and two bound states) present any difficulties since the two signals are located very
obeys the modified Bloch equations. The amplitude of each far from each other (δ ϭ 190 ppm) and a wide pulse, in
resonance as a function of the mixing time is given by the the frequency domain, can be applied. However, our system
following matrix equation [Equation (1)], where Ii0 is the ini- cannot be studied over a large range of temperatures. Above
tial magnetization of the nucleus at τm ϭ 0, and Iieq is the 248 K the exchange proceeded too rapidly, and the signal-
magnetization at this site for infinite relaxation time.
to-noise ratio decreased significantly. At too low a temper-
978
Eur. J. Org. Chem. 2003, 973Ϫ984

Xe@cryptophane Complexes with C2 Symmetry
FULL PAPER
ature cryptophanes might precipitate or the solvent crystal-
lize. Temperatures ranging from 248 K to 228 K were found
to be suitable for these experiments using 1,1,2,2-[D2]te-
trachloroethane as solvent. We made sure that even at
228 K the solvent did not crystallize and no precipitate
was observed.
Figure 5 shows the redistribution of the magnetization at
238 K as a function of the mixing time when both free and
bound xenon atoms have been selectively irradiated. It took
about 6.5 h to obtain rate constants at a given temperature
using 32 scans for each acquisition for the 12 mixing time
values. The intensities of each resonance were then meas-
ured and the rate constants for decomplexation were deter-
mined by fitting the experimental data as a function of the
mixing time τm (Figure 6).
Rate constants decrease by a factor of five when the tem-
perature varies from 248 K to 228 K (Table 2). The corres-
ponding activation energies were obtained from the Arrhen-
ius
Figure 6. Evolution of the magnetization as a function of the mix-
ing time after selective irradiation of free and encapsulated xenon;
each panel represents a particular site and its evolution with τm
(only mixing times ranging from 0 to 40 ms are shown); experi-
mental data are represented by dots and the solid lines represent
the best fit of the data
plot, lnk vs. 1/T, to give Ea ϭ 35.5 kJ molϪ1 and Ea
ϭ
33.1 kJ molϪ1 for cryptophane-A and cryptophane-223, re-
spectively (Figure 7, Table 4). The value obtained for crypt-
ophane-A is close to that previously reported, Ea ϭ 37.5 kJ
molϪ1 [9]
A lower value of Ea was found for cryptophane-
.
223, consistent with an increase of the size of the cavity,
which causes a decrease in the hostϪguest interactions.
However, it should be noticed that the fit of the experi-
mental data required a set of 14 independent parameters
(nine for the initial magnetizations, three for the magnetiza-
tion at equilibrium and two for rate constants). Although
the use of a sample containing two cryptophanes has some
advantages, it also makes the treatment of the data more
complicated.
Table 2. Rate constants for decomplexation of Xe@cryptophane-
A (k1,3, M03/M01 ϭ 1.1) and Xe1 (k2,3, M03/M02 ϭ 1.8) obtained from
variable temperature 129Xe 1D-EXSY experiments in 1,1,2,2-
[D2]tetrachloroethane (Experiment 1)
T (K)
248
243
238
233
228
k1,3 (sϪ1
k2,3 (sϪ1
)
)
97.0 Ϯ 5 66.0 Ϯ 3 44.0 Ϯ 3 29.5 Ϯ 2 21.6 Ϯ 2
104.9 Ϯ 10 70.7 Ϯ 5 51.6 Ϯ5 35.8 Ϯ 2 24.9 Ϯ 2
In a second experiment (Experiment 2), a solution con-
taining exclusively host 1 and an excess of xenon was used.
1D-EXSY experiments were then performed at different
temperatures ranging from 238 K to 225 K. Similar experi-
ments were run with a sample containing host 2 and an
excess of xenon. Typical fits of the experimental data ob-
tained with 1 at 233 K are shown in Figure 8. In this case,
the fit of the experimental data requires only seven inde-
pendent parameters (four for the initial magnetization, two
for the magnetization at equilibrium and one for the rate
Figure 5. Evolution of the magnetization of the three sites as a
function of the mixing time at 238 K (1D-EXSY): a) after selective
excitation of the peak corresponding to free xenon; b) after select-
ive excitation of the two peaks corresponding to the Xe@crypto-
phane complexes
Eur. J. Org. Chem. 2003, 973Ϫ984
979

T. Brotin, J.-P. Dutasta
FULL PAPER
Table 3. Rate constants for decomplexation of Xe1 (c ഠ 0.06 ,
[
129Xe]/[129Xe1] ഠ 1), and Xe 2 (c ഠ 0.06 , [129Xe]/[129Xe2] ഠ 1),
obtained from 129Xe 1D-EXSY experiments at several temperatures
in 1,1,2,2-[D2]tetrachloroethane (Experiment 2).
T (K)
243
238
233
228
225
k (sϪ1) 1 64.7 Ϯ 5 39.1 Ϯ 2 26.9 Ϯ 2 18.5 Ϯ 1 16.5 Ϯ 2
k (sϪ1) 2 69.8 Ϯ 6 48.9 Ϯ 2 33.1 Ϯ 2 21.7 Ϯ 2 18.6 Ϯ 1
Figure 7. Arrhenius plot of ln(k1,3) and ln(k2,3) vs. 1/T; values of
k1,3 and k2,3 obtained from 129Xe 1D-EXSY experiments at differ-
ent temperatures ranging from 248 K to 228 K for cryptophane-A
(o) and 1 (᭹)
constant) since only two exchanging sites are present in so-
lution and the fitting curves are now represented by four
panels.
Figure 9. Arrhenius plot of lnk vs. 1/T obtained from 129Xe 1D-
EXSY experiments performed at different temperatures in 1,1,2,2-
[D2]tetrachloroethane (k stands for the decomplexation rate con-
stant); a) 1 (238 to 225 K), b) 2 (243 to 225 K)
The decomplexation activation enthalpy ∆H‡ and en-
tropy ∆S‡ for cryptophane-A and compounds 1 and 2 were
then obtained by plotting ln(k/T) vs. 1/T according to the
Eyring equation [Equation (3)], where T is the absolute tem-
perature and R is the gas constant (Table 4).
Figure 8. Evolution of the magnetizations as a function of mixing
time, after selective irradiation of one of the two sites of a solution
containing cryptophane-223 (1) and an excess of xenon at 233 K;
each panel represents a particular site and its evolution with τm
(only mixing times ranging from 10 µs to 100 ms are shown); ex-
perimental data are represented by dots and the solid lines repres-
ent the best fit of the data
(3)
Large negative ∆S‡ values (ഠ Ϫ72 J molϪ1
K
Ϫ1) were
found for the new cryptophanes. As already reported for
The rate constants for decomplexation obtained for cryptophane-A this suggests that the activated Xe1 and Xe2
cryptophanes 1 and 2 at several temperatures are reported complexes are conformationally more strained than the
in Table 3. The decomplexation activation energies were ob- complexes in their ground states. The values of the Gibbs
tained from the Arrhenius plots (Figure 9). The new value activation energy, ∆G‡ ഠ 50 kJ molϪ1 at 238 K, are similar
Ea ϭ 35.1 kJ molϪ1 for Xe1 is close to that previously calcu- for all cryptophanes, within the limit of experimental error.
lated in Experiment 1 (Table 4), suggesting that the mixing The interpretation of these results is not straightforward,
of two cryptophanes in the same sample has no con- as molecules 1 and 2 do not have the same symmetry as
sequences for the result, even though we believe that Ex- cryptophane-A. The introduction of two linkers with differ-
periment 2 gives more accurate data. The value for the de- ent length decreases the symmetry of the new cryptophanes
complexation activation energy Ea ϭ 34.8 kJ molϪ1 for Xe2 1 and 2 (C2 symmetry instead of D3 symmetry for crypto-
is very close to that obtained for Xe1.
phane-A), and therefore creates two portals with different
980
Eur. J. Org. Chem. 2003, 973Ϫ984

Xe@cryptophane Complexes with C2 Symmetry
FULL PAPER
Table 4. Calculated decomplexation activation energy for cryptophane-A, 1 and 2 obtained from the Arrhenius plot ln(k) vs. 1/T; the
decomplexation activation enthalpy ∆H‡ and entropy ∆S‡ have been calculated from an Eyring plot ln(k/T) vs. 1/T; the Gibbs activation
energy of decomplexation at T ϭ 238 K is also given
Ea (kJ/mol)
∆H‡ (kJ/mol)
∆S‡ (J/mol K)
∆G‡ (kJ/mol)
Cryptophane-A[a]
Cryptophane-A (Exp. 1)
37.5 Ϯ 2.0
35.5 Ϯ 2.0
(r2 ϭ 0.993)
33.1 Ϯ 3.0
(r2 ϭ 0.999)
35.1 Ϯ 2.0
(r2 ϭ 0.980)
34.8 Ϯ 2.0
(r2 ϭ 0.998)
35.5 Ϯ 2.0
33.6 Ϯ 2.0
(r2 ϭ 0.993)
31.3 Ϯ 3.0
(r2 ϭ 0.999)
33.2 Ϯ 2.0
(r2 ϭ 0.978)
32.8 Ϯ 2.0
(r2 ϭ 0.998)
Ϫ60.0 Ϯ 5
Ϫ70.5 Ϯ 10.0
49.8 Ϯ 2.0
50.4 Ϯ 2.0
Cryptophane 1 (Exp. 1)
Cryptophane 1 (Exp. 2)
Cryptophane 2 (Exp. 2)
Ϫ79.0 Ϯ 10.0
Ϫ72.8 Ϯ 10.0
Ϫ71.7 Ϯ 10.0
50.0 Ϯ 2.0
50.5 Ϯ 2.0
49.9 Ϯ 2.0
[a]
From ref.[9]
sizes. This may therefore complicate the interpretation of Symmetrical cryptophanes are thus good models for investi-
the experimental data since xenon may now use either of gating complexation dynamics and for correlating the phys-
two different windows to leave molecules 1 and 2.
ical properties of the encapsulated xenon with its environ-
Slight structural modifications, such as the addition of ment. Efforts to design new cryptophane hosts with modi-
one or two methylene groups in the linkers, induce a sub- fied structures are currently underway, with particular ap-
stantial change of the 129Xe chemical shift, and have im- plication to the design of biosensors for magnetic
portant consequences on the stability of the Xe@crypto- resonance imaging.[19]
phane complex. However, this structural change seems to
have only a small effect on the activation energies Ea. Addi-
tional support from molecular modeling would be helpful
in exploring the possible ways in which the xenon guest may
Experimental Section
be extruded from the cryptophane cavity, and may give new
insights leading to better interpretation of these findings.
General Remarks: 1,2-dibromoethane, 1,3-dibromopropane, 1,4-di-
bromobutane, allyl bromide and pyridinium p-toluenesulfonate
(PPTS) were purchased from Acros Organics, and were used with-
out further purification. Cesium carbonate and potassium carbon-
ate were dried under vacuum before use. Solvents were distilled
prior to use (DMF from calcium hydride under reduced pressure,
acetone and CH2Cl2 from calcium chloride, THF from benzo-
phenone ketyl ether). Column chromatographic separations were
carried out over Merck silica gel 60 (0.040Ϫ0.063 mm). Analytical
thin layer chromatography (TLC) was performed on MERCK
silica-gel TLC plates F-254. Melting points were measured on a
PerkinϪElmer DSC7 microcalorimeter. Unprotected 7, 8, 12 and
13,[12,13] and protected 10 and 15,[9] halogenated vanillyl alcohol
derivatives, and monosubstituted cyclotriveratrylene 17,[9] were
synthesized using literature procedures reported by our group.
Conclusion
We have prepared the new cryptophanes 1Ϫ3 with C2
symmetry where the two cyclotriveratrylene units are con-
nected with different alkyl chains. These compounds have
been isolated as the chiral anti isomer, in moderate yields.
The synthetic route follows a stepwise sequence of reac-
tions, which proved to be very useful for preparing new se-
lectively protected cryptophanes, key compounds for the
elaboration of large supramolecular structures.[22] We have
demonstrated that molecules 1 and 2 efficiently complex
xenon in organic solution, with binding constants slightly
lower than that measured with cryptophane-A, according
to the increased size of the molecular cavities. Under the
same experimental conditions, the affinity of xenon for 3 is
considerably less than for 1 and 2. No signal was detected
NMR Measurements: 129Xe NMR experiments were carried out at
138.4 MHz on a Varian Unityϩ spectrometer using a 10-mm
double resonance probe. The 129Xe chemical shift is expressed in
ppm with respect to xenon gas extrapolated to zero pressure. For
competition experiments equimolar amounts of cryptophane-A
(110 mg) and cryptophane-223 (1) (112 mg) were dissolved in
for encapsulated xenon, suggesting fast-exchange condi- 1,1,2,2-[D2]tetrachloroethane. The solvent was evaporated under
reduced pressure to remove any bound substrate (mostly CHCl3).
1,1,2,2-[D2]Tetrachloroethane (3.5 mL) was then added and the so-
lution was poured into a 10 mm NMR tube (c ഠ 0.035  for each
cryptophane). Xenon gas (129Xe natural abundance 26.4%) was
then bubbled through the solution. The same method was applied
to prepare the NMR tube containing equimolar amounts of
cryptophane-223 (1) (150 mg) and cryptophane-233 (2) (150 mg) in
1,1,2,2-[D2]tetrachloroethane (3.7 mL, c ഠ 0.045  for each crypto-
phane), and the sample containing compound 3 (150 mg) in
1,1,2,2-[D2]tetrachloroethane (3.5 mL, c ഠ 0.046 ). 1H and 13C
tions on the NMR timescale. In addition, a linear relation-
ship was observed between the measured binding constant
K and the volume of the inner cavity of the cryptophanes.
Measurement of the decomplexation activation energy
and the associated ∆H‡ and ∆S‡ parameters from 1D-
EXSY experiments gave values in good agreement with
those previously reported for cryptophane-A. Once more,
the high sensitivity of the 129Xe chemical shift proved to
be a powerful tool for discriminating between the slightly
modified environments provided by aromatic cavities. C2- NMR spectra were run on a Varian Unityϩ 500 spectrometer (s:
Eur. J. Org. Chem. 2003, 973Ϫ984 981

T. Brotin, J.-P. Dutasta
FULL PAPER
singlet, d: doublet, t: triplet, q: quadruplet, quint: quintuplet, m: 112.98, 110.79, 67.82, 65.11, 55.82 (1 C, OCH3), 30.06, 29.99, 6.49
multiplet). The bridging methylene carbon of the cyclotriveratry- (1 C, CH2I) ppm. C12H17IO3 (336.17): calcd. C 42.88, H 5.10;
lene moieties were labeled Ca,e, and the corresponding protons were found C 42.70, H 5.03.
labeled Ha (axial) and He (equatorial).
2-[4-(3-Bromopropoxy)-3-methoxybenzyloxy]tetrahydropyran (11):
Complexation Dynamics as a Function of Temperature: The NMR
sample containing equimolar amounts of cryptophane-A (110 mg)
and cryptophane-223 (112 mg) was prepared as described above. It
was then frozen in liquid nitrogen and sealed under vacuum (c ഠ
0.035  for each cryptophane). For 129Xe 1D-EXSY NMR experi-
ments, the sample was cooled to 248 K and kept at this temperature
A solution of PPTS (0.55 g, 2.18 mmol) in CH2Cl2 (10 mL) was
added in one portion to a mixture of 8 (6.0 g, 21.8 mmol) and dihy-
dropyran (2.65 g, 31.5 mmol) in THF (70 mL). The solution was
stirred overnight at room temperature under an argon atmosphere.
The solvent was removed to leave a residue which was extracted
with diethyl ether. The combined organic layers were washed twice
for about 2 h before starting experiments. The mixing times ranged with brine, dried over Na2SO4 and the solvents evaporated to leave
from 0.01 ms to 5 s (10 or 12 values of τm were used for each tem- an oily residue. Chromatography on silica gel (diethyl ether/pent-
perature). A Gaussian pulse with duration of 139 µs was used for ane, 50:50) yielded compound 11 (7.0 g, 89%) as a white solid, m.p.
1
the 180° selective pulse. A relaxation delay of 30 s was used for
each increment. Each spectrum was recorded with 32 scans for a
given mixing time. The duration of the 90° pulse was 25 µs. This
required about 6.5 h to complete an experiment. After each change
37 °C. H NMR (500 MHz, CDCl3, 20 °C): δ ϭ 6.90Ϫ6.85 (m, 3
2 2
H, Ar), 4.70 (d, J ϭ 11.5 Hz, 1 H), 4.67 (m, 1 H), 4.42 (d, J ϭ
11.5 Hz, 1 H), 4.13 (t, 3J ϭ 6.0 Hz, 2 H, OCH2), 3.91 (m, 1 H),
3
3.85 (s, 3 H, OCH3), 3.61 (t, J ϭ 6.0 Hz, 2 H, CH2Br), 3.53 (m, 1
of temperature the sample was left for 30 min before starting an- H), 2.34 (quint, 3J ϭ 6.0 Hz, 2 H, CH2), 1.90Ϫ1.50 (m, 6 H, THP)
other experiment. In order to follow the evolution of rate constants
with temperature and thus to obtain a better fit, a different set of
mixing times was chosen for each temperature. Prior to the Fourier
transform an exponential apodization (lb ϭ 20) was applied. The
intensity of each peak was then measured for a given vertical scale
value, which was kept the same for all experiments. The NMR
samples containing only molecule 1 (200 mg, c ഠ 0.06 ) or molec-
ule 2 (200 mg, c ഠ 0.06 ) in 1,1,2,2-[D2]tetrachloroethane (3.5 mL)
and in the presence of an excess of xenon were prepared as de-
scribed above. A Gaussian pulse with duration of 278 µs was used
for the 180° selective pulse. All other parameters were the same as
those described above.
ppm. 13C NMR (125.67 MHz, CDCl3, 20 °C): δ ϭ 149.42, 147.53,
131.31, 120.46, 113.32, 111.77, 97.47, 68.64, 66.64, 62.19, 55.80 (1
C, OCH3), 32.23, 30.50, 30.10, 25.35, 19.39 ppm. C16H23BrO4
(359.26): calcd. C 53.49, H 6.45; found C 53.40, H 6.71.
2-[4-(3-Iodopropoxy)-3-methoxybenzyloxy]tetrahydropyran (16): So-
dium iodide (12.5 g, 83.5 mmol) was added in one portion to a
stirred solution of 11 (6.0 g, 16.7 mmol) in dry acetone. The solu-
tion was refluxed overnight under an argon atmosphere. After cool-
ing to room temperature, the solvent was removed by rotary evap-
oration to leave a residue which was extracted with diethyl ether.
The combined organic layers were washed with brine, dried over
Na2SO4 and the solvents evaporated to leave a residue. Chromato-
graphy on silica gel (diethyl ether/pentane, 50:50) yielded 16 (6.44 g,
95%) as a white solid, m.p. 50 °C. 1H NMR (500 MHz, CDCl3, 20
(4-Bromobutoxy-3-methoxyphenyl)methanol (9): 1,4-Dibromobut-
ane (12.63 g, 58.3 mmol) was added under argon to a stirred solu-
tion of
4
(3.0 g, 19.5 mmol), potassium carbonate (2.7 g, °C): δ ϭ 6.90Ϫ6.85 (m, 3 H, Ar), 4.70 (d, 2J ϭ 11.5 Hz, 1 H), 4.67
2 3
19.5 mmol) in freshly distilled acetone (50 mL). The mixture was
refluxed overnight under an argon atmosphere. The solvent was
removed under reduced pressure and the residue was extracted with
(m, 1 H), 4.42 (d, J ϭ 11.5 Hz, 1 H), 4.06 (t, J ϭ 6.0 Hz, 2 H,
OCH2), 3.91 (m, 1 H), 3.85 (s, 3 H, OCH3), 3.53 (m, 1 H), 3.37 (t,
3J ϭ 6.0 Hz, 2 H, CH2I), 2.29 (quint, 3J ϭ 6.5 Hz, 2 H, CH2),
ethyl acetate. The combined organic layers were washed three times 1.90Ϫ1.50 (m, 6 H, THP) ppm. 13C NMR (125.67 MHz, CDCl3,
with brine, dried over Na2SO4 and the solvent was evaporated to
leave a residue. Column chromatography (silica gel, eluent: EtOAc/
pentane, 70:30) yielded 9 as a white solid (4.2 g, 75%), m.p. 58 °C.
1H NMR (500 MHz, CDCl3, 20 °C): δ ϭ 6.91Ϫ6.81 (m, 3 H, Ar),
20 °C): δ ϭ 149.43, 147.51, 131.32, 120.46, 113.38, 111.79, 97.46,
68.59, 68.45, 62.19, 55.82 (1 C, OCH3), 32.85, 30.50, 25.36, 19.39,
2.63 (1 C, CH2I) ppm. C16H23IO4 (406.26): calcd. C 47.30, H 5.71;
found C 47.14, H 5.61.
3
4.59 (s, 2 H, CH2OH), 4.02 (t, J ϭ 6.0 Hz, 2 H, CH2), 3.85 (s, 3
12-[3-(4-Hydroxymethyl-2-methoxyphenoxy)propoxy]-3,8,13-tri-
methoxy-10,15-dihydro-5H-tribenzo[a,d,g]cyclononene-2,7-diol (18).
General Procedure A: Compound 13 (0.63 g, 1.96 mmol) was added
in one portion to a stirred solution of 6 (0.8 g, 1.96 mmol) and
cesium carbonate (1.28 g, 3.93 mmol) in dry DMF (35 mL). The
mixture was stirred overnight at 80 °C under an argon atmosphere.
After cooling to room temperature, the mixture was poured into
water and the product extracted with ethyl acetate. The combined
organic layers were washed several times with brine and then dried
over Na2SO4. The solvent was removed to leave a residue which
was chromatographed on silica gel (eluent: EtOAc). The second
spot detected by TLC was collected and corresponds to the ex-
pected derivative 18 (0.38 g, 32%) which was isolated as a glassy
solid. 1H NMR (500 MHz, CDCl3, 20 °C): δ ϭ 6.91Ϫ6.73 (m, 9
H, OCH3), 3.48 (t, 3J ϭ 6.0 Hz, 2 H, CH2Br), 2.04 (quint, 3J ϭ
3
7.0 Hz, 2 H, CH2), 1.98 (quint, J ϭ 7.0 Hz, 2 H, CH2), 1.64 (s, 1
H, OH) ppm. 13C NMR (125.67 MHz, CDCl3, 20 °C): δ ϭ 149.44,
147.66, 133.87, 119.27, 112.96, 110.78, 68.00, 65.08, 55.80 (1 C,
OCH3), 33.48, 29.34, 27.73. C12H17BrO3 (289.17): calcd. C 49.84,
H 5.93; found C 49.79, H 5.93.
(4-Iodobutoxy-3-methoxyphenyl)methanol (14): Sodium iodide
(7.8 g, 54 mmol) was added in one portion to a stirred solution of 9
(3.0 g, 10.3 mmol) in dry acetone (60 mL). The solution was stirred
overnight under an argon atmosphere. After cooling, the solvent
was removed to leave a residue which was extracted with ethyl ace-
tate. The combined organic layers were washed three times with
brine and dried over Na2SO4. Chromatography on silica gel
(EtOAc/pentane, 70:30) yielded 14 as a white solid (3.3 g, 95%), H, Ar), 5.40 (s, 1 H, OH), 5.37 (s, 1 H, OH), 4.715 (d, 2J ϭ 14.0 Hz,
m.p. 66 °C. 1H NMR (500 MHz, CDCl3, 20 °C): δ ϭ 6.91Ϫ6.81 2 H, CHa), 4.70 (d, 2J ϭ 13.5 Hz, 2 H, CHa), 4.69 (d, 2J ϭ 13.5 Hz,
(m, 3 H, Ar), 4.60 (d, 3J ϭ 5.5 Hz, 2 H, CH2OH), 4.02 (t, 3J ϭ
2 H, CHa), 4.57 (d, J ϭ 6.0 Hz, 2 H, CH2), 4.28Ϫ4.18 (m, 2 H,
3
6.5 Hz, 2 H, OCH2), 3.85 (s, 3 H, OCH3), 3.25 (t, J ϭ 6.5 Hz, 2 OCH2), 4.16 (t, 3J ϭ 6.5 Hz, 2 H, OCH2), 3.84 (s, 3 H, OCH3),
3
3
3
H, CH2I), 2.02 (quint, J ϭ 7.0 Hz, 2 H, CH2), 1.92 (quint, J ϭ 3.78 (s, 3 H, OCH3), 3.74 (s, 3 H, OCH3), 3.72 (s, 3 H, OCH3),
7.0 Hz, 2 H, CH2), 1.58 (t, 3J ϭ 6.5 Hz, 1 H, OH) ppm. 13C NMR
3.49 (d, J ϭ 13.5 Hz, 6 H, CHe), 2.27 (quint, J ϭ 6.5 Hz, 2 H,
2
3
(125.67 MHz, CDCl3, 20 °C): δ ϭ 149.46, 147.68, 133.86, 119.28, CH2), 1.61 (t, 3J ϭ 6.0 Hz, 1 H, OH) ppm. 13C NMR (125.67 MHz,
982
Eur. J. Org. Chem. 2003, 973Ϫ984

Xe@cryptophane Complexes with C2 Symmetry
CDCl3, 20 °C): δ ϭ 149.48, 148.17, 147.80, 146.76, 145.21, 145.17, 3 H, OCH3), 3.72 (s, 3 H, OCH3), 3.65 (s, 3 H, OCH3), 3.58Ϫ3.46
FULL PAPER
144.05, 143.98, 133.81, 132.47, 132.28, 132.17, 131.82, 131.31,
131.15, 119.40, 115.50 (2 C), 115.30, 113.52, 113.10, 112.18, 112.13,
(m, 5 H), 2.28 (m, 4 H, CH2), 1.90Ϫ1.50 (m, 13 H, THPϩOH)
ppm. 13C NMR (125.67 MHz, CDCl3, 20 °C): δ ϭ 149.66, 149.32
110.77, 65.93, 65.71, 65.23, 56.06 (1 C, OCH3), 56.01 (1 C, OCH3), (2 C), 148.46, 148.15, 148.09, 147.77 (2 C), 147.42, 146.85 (2 C),
55.88 (1 C, OCH3), 55.67 (1 C, OCH3), 36.35 (1 C, CHa,e), 36.19 146.73, 134.52, 133.05, 132.12, 132.10, 131.84, 131.83, 131.71,
(2 C, CHa,e), 29.05 (1 C, CH2) ppm. HRMS (LSIMS): exact mass 130.99, 130.97, 120.53 (2 C), 119.30, 116.70 (2 C), 115.10, 113.81,
calcd. for C35H37O9 [M Ϫ H]Ϫ 601.2438, found 601.2432.
113.71 113.53 (2 C), 112.89 (2 C), 111.72 (2 C), 110.88, 97.49 (2
C), 68.71 (2 C), 68.09 (2 C), 67.75 (2 C), 65.87, 65.59, 65.14, 62.26
(2 C), 56.05 (3 C, 3 ϫ OCH3), 55.80 (2 C, 2 ϫ OCH3), 55.69 (1 C,
OCH3), 36.36 (3 C, CHa,e), 30.56 (2 C), 29.07 (2 C, CH2), 25.42
(2 C), 19.44 (2 C) ppm. HRMS (LSIMS): exact mass calcd. for
C66H80NaO17 [M ϩ Na]ϩ 1167.5293, found 1167.5289.
12-[3-(4-Hydroxymethyl-2-methoxyphenoxy)butoxy]-3,8,13-tri-
methoxy-10,15-dihydro-5H-tribenzo[a,d,g]cyclononene-2,7-diol (19):
Using general procedure A, 19 (0.4 g, 33%) was obtained from
cyclotriveratrylene 6 (0.8 g, 1.96 mmol), compound 14 (0.65 g,
1.93 mmol) and cesium carbonate (1.28 g, 3.92 mmol) in DMF
1
(45 mL). H NMR (500 MHz, CDCl3, 20 °C): δ ϭ 6.87Ϫ6.74 (m,
{3-Methoxy-4-[4-(3,8,13-trimethoxy-7,12-bis{2-[2-methoxy-4-(tetra-
hydropyran-2-yloxymethyl)phenoxy]ethoxy}-10,15-dihydro-5H-
tribenzo[a,d,g]cyclononen-2-yloxy)butoxy]phenyl}methanol (22): Us-
ing general procedure A, compound 22 (0.59 g, 79%) was obtained
from 19 (0.40 g, 0.65 mmol), 15 (0.61 g, 1.50 mmol) and cesium
carbonate (0.85 g, 2.6 mmol) in DMF (25 mL). 1H NMR
(500 MHz, CDCl3, 20 °C): δ ϭ 7.01Ϫ6.78 (m, 15 H, Ar), 4.72 (d,
9 H, Ar), 5.45 (s, 1 H, OH), 5.40 (s, 1 H, OH), 4.72 (d, 2J ϭ
13.5 Hz, 1 H, CHa), 4.70 (d, 2J ϭ 13.5 Hz, 1 H, CHa), 4.69 (d, 2J ϭ
13.5 Hz, 1 H, CHa), 4.59 (d, 3J ϭ 6.0 Hz, 2 H, CH2), 4.10Ϫ3.96 (m,
4 H, OCH2), 3.84 (s, 3 H, OCH3), 3.79 (s, 3 H, OCH3), 3.73 (s, 3
2
H, OCH3), 3.64 (s, 3 H, OCH3), 3.48 (d, J ϭ 13.5 Hz, 3 H, CHe),
1.98 (m, 4 H, CH2), 1.61 (t, 3J ϭ 6.0 Hz, 1 H, OH) ppm. 13C NMR
(125.67 MHz, CDCl3, 20 °C): δ ϭ 149.49, 148.25, 147.96, 146.94,
145.22, 145.17, 144.08 (2 C), 133.57, 132.56, 132.33, 132.07, 131.66,
131.28, 131.12, 119.41, 115.59, 115.50, 115.06, 113.53, 112.79,
112.18, 112.08, 110.82, 68.80, 68.61, 65.36, 56.10 (1 C, OCH3),
56.06 (1 C, OCH3), 55.91 (1 C, OCH3), 55.63 (1 C, OCH3), 36.44
(1 C, CHa,e), 36.21 (2 C, CHa,e), 25.98 (1 C, CH2), 25.90 (1 C, CH2)
ppm. HRMS (LSIMS): exact mass calcd. for C36H39O9 [M Ϫ H]Ϫ
615.2594, found 615.2595.
2
2J ϭ 13.5 Hz, 3 H, CHa), 4.70 (d, J ϭ 11.5 Hz, 2 H), 4.66 (m, 2
3
2
H), 4.58 (d, J ϭ 6.0 Hz, 2 H, CH2OH), 4.42 (d, J ϭ 11.5 Hz, 2
H), 4.34 (m, 8 H, CH2), 4.10Ϫ3.96 (m, 4 H, CH2), 3.90 (m, 2 H),
3.84 (s, 6 H, 2 ϫ OCH3), 3.72 (s, 3 H, OCH3), 3.70 (s, 9 H, 3 ϫ
OCH3), 3.56Ϫ3.48 (m, 5 H), 1.98 (m, 4 H, CH2), 1.90Ϫ1.50 (m, 13
H, THPϩOH) ppm. 13 C NMR (125.67 MHz, CDCl3, 20 °C): δ ϭ
149.42 (2 C), 149.39, 148.40, 148.30, 148.22, 147.82, 147.49 (2 C),
147.00, 146.78 (2 C), 133.76, 132.96, 132.84, 131.99, 131.91, 131.88,
131.53, 131.51, 131.49, 120.49 (2 C), 119.31, 116.53, 116.50, 114.96,
113.73, 113.53 (3 C), 113.47, 112.78, 111.85 (2 C), 110.72, 97.54 (2
C), 68.66 (3 C), 68.54, 68.12 (2 C), 67.59 (2 C), 65.18, 62.24 (2 C),
56.16 (1 C, OCH3), 56.08 (1 C, OCH3), 56.04 (1 C, OCH3), 55.81
(2 C, 2 ϫ OCH3), 55.64 (1 C, OCH3), 36.40 (1 C, CHa,e), 36.33 (2
C, CHa,e), 30.53 (2 C), 25.93 (1 C, CH2), 25.83 (1 C, CH2), 25.38
(2 C), 19.42 (2 C) ppm. HRMS (LSIMS): exact mass calcd. for
C66H80O17 [M]ϩ 1144.5396, found 1144.5375.
{3-Methoxy-4-[3-(3,8,13-trimethoxy-7,12-bis{2-[2-methoxy-4-(tetra-
hydropyran-2-yloxymethyl)phenoxy]ethoxy}-10,15-dihydro-5H-
tribenzo[a,d,g]cyclononen-2-yloxy)propoxy]phenyl}methanol
(21):
Using general procedure A, compound 21 (0.5 g, 72%) was ob-
tained from 18 (0.37 g, 0.62 mmol), 15 (0.58 g, 1.48 mmol) and ce-
sium carbonate (0.81 g, 2.48 mmol) in DMF (20 mL). 1H NMR
(500 MHz, CDCl3, 20 °C): δ ϭ 7.01Ϫ6.78 (m, 15 H, Ar), 4.72 (d,
2
2J ϭ 13.0 Hz, 3 H, CHa), 4.70 (d, J ϭ 11.5 Hz, 2 H), 4.66 (m, 2
3
2
H), 4.57 (d, J ϭ 6.0 Hz, 2 H, CH2OH), 4.42 (d, J ϭ 11.5 Hz, 2
Cryptophane-223 (1). General Procedure B: In a 1-L rotary evapo-
rator flask, compound 21 (0.6 g, 0.53 mmol) was dissolved in chlo-
roform (6 mL). Then formic acid (600 mL) was added in one por-
tion and the solution was heated at 55 °C for 2 h 30 min with slow
rotation. The solvent was evaporated under reduced pressure and
chloroform was added to remove residual formic acid by azeotropic
distillation. The cryptophane was purified by column chromato-
graphy on silica gel (CH2Cl2/acetone, 90:10) and washed with a
few milliliters of diethyl ether on a frit. A second run of column
chromatography (CH2Cl2/acetone, 90:10) yielded pure crypto-
phane-223 (0.17 g, 35%); decomp. above 300 °C. 1H NMR
(500 MHz, CDCl3, 20 °C): δ ϭ 6.77 (s, 2 H, Ar), 6.73 (s, 2 H, Ar),
6.67 (s, 2 H, Ar), 6.66 (s, 2 H, Ar), 6.64 (s, 2 H, Ar), 6.55 (s, 2 H,
3
H), 4.33 (m, 8 H, CH2), 4.26Ϫ4.16 (m, 2 H, OCH2), 4.15 (t, J ϭ
6.0 Hz, 2 H, OCH2), 3.90 (m, 2 H), 3.84 (s, 6 H, 2 ϫ OCH3), 3.77
(s, 3 H, OCH3), 3.72 (s, 3 H, OCH3), 3.69 (s, 6 H, 2 ϫ OCH3),
3.56Ϫ3.48 (m, 5 H), 2.27 (p, 3J ϭ 6.0 Hz, 2 H, CH2), 1.90Ϫ1.50
(m, 13 H, THP ϩ OH) ppm. 13C NMR (125.67 MHz, CDCl3, 20
°C): δ ϭ 149.48, 149.43 (2 C), 148.42, 148.37, 148.21, 147.74,
147.50 (2 C), 146.85, 146.77, 146.76, 133.94, 133.00, 132.91, 132.12,
131.90, 131.80, 131.66, 131.53 (2 C), 120.52 (2 C), 119.34, 116.51,
116.44, 115.20, 113.74, 113.70, 113.48 (3 C), 113.03, 111.84 (2 C),
110.73, 97.56 (2 C), 68.69 (2 C), 68.10 (2 C), 67.59 (2 C), 65.85,
65.65, 65.20, 62.28 (2 C), 56.16 (1 C, OCH3), 56.08 (1 C, OCH3),
56.04 (1 C, OCH3), 55.81 (2 C, 2 ϫ OCH3), 55.64 (1 C, OCH3),
36.36 (3 C), 30.56 (2 C), 29.09 (1 C, CH2), 25.41 (2 C), 19.44 (2
C) ppm. HRMS (LSIMS): exact mass calcd. for C65H78O17 [M]ϩ
1130.5239, found 1130.5286.
2
2
Ar), 4.61 (d, J ϭ 14.0 Hz, 4 H, CHa), 4.59 (d, J ϭ 14.0 Hz, 2 H,
CHa), 4.31Ϫ4.22 (m, 2 H, OCH2), 4.16Ϫ3.98 (m, 8 H, OCH2),
3.89Ϫ3.82 (m, 2 H, OCH2), 3.81 (s, 6 H, 2 ϫ OCH3), 3.79 (s, 6 H,
2
{3-Methoxy-4-[2-(3,8,13-trimethoxy-7,12-bis{3-[2-methoxy-4-(tetra-
hydropyran-2-yloxymethyl)phenoxy]propoxy}-10,15-dihydro-5H-
tribenzo[a,d,g]cyclononen-2-yloxy)ethoxy]phenyl}methanol (20): Us-
ing general procedure A, compound 20 (0.75 g, 74%) was obtained
2 ϫ OCH3), 3.77 (s, 6 H, 2 ϫ OCH3), 3.41 (d, J ϭ 14.0 Hz, 2 H,
2
CHe), 3.39 (d, J ϭ 14.0 Hz, 4 H, CHe), 2.29 (m, 2 H, CH2) ppm.
13C NMR (125.67 MHz, CDCl3, 20 °C): δ ϭ 149.45 (2 C), 149.37
(2 C), 147.44 (2 C), 147.31 (2 C), 146.68 (2 C), 146.52 (2 C), 133.97
from 17 (0.52 g, 0.88 mmol), 16 (0.86 g, 2.12 mmol) and cesium (2 C), 133.83 (2 C), 132.43 (2 C), 131.39 (2 C), 131.25 (2 C), 130.48
carbonate (1.15 g, 3.5 mmol) in DMF (30 mL). 1H NMR (2 C), 120.43 (2 C), 120.33 (2 C), 114.01 (2 C), 113.46 (2 C), 112.21
(500 MHz, CDCl3, 20 °C): δ ϭ 6.98Ϫ6.77 (m, 15 H, Ar), 4.71 (d,
(2 C), 112.08 (2 C), 69.63 (2 C), 68.94 (2 C), 63.66 (2 C), 56.22 (2
C, 2 ϫ OCH3), 55.54 (2 C, 2 ϫ OCH3), 55.43 (2 C, 2 ϫ OCH3),
2
2J ϭ 13.5 Hz, 3 H, CHa), 4.69 (d, J ϭ 11.5 Hz, 2 H), 4.65 (m, 2
3
2
H), 4.58 (d, J ϭ 6.0 Hz, 2 H, CH2OH), 4.40 (d, J ϭ 11.5 Hz, 2 36.41 (2 C, CHa,e), 36.05 (2 C, CHa,e), 35.99 (2 C, CHa,e), 29.84 (1
H), 4.38Ϫ4.29 (m, 4 H, OCH2), 4.26Ϫ4.12 (m, 8 H, OCH2), 3.90
C, CH2) ppm. HRMS (LSIMS): exact mass calcd. for C55H56O12
(m, 2 H), 3.82 (s, 6 H, 2 ϫ OCH3), 3.73 (s, 3 H, OCH3), 3.725 (s, [M]ϩ, 908.377, found 908.3780.
Eur. J. Org. Chem. 2003, 973Ϫ984
983

T. Brotin, J.-P. Dutasta
FULL PAPER
[1]
[2]
[3]
[4]
[5]
[6]
[7]
Cryptophane-233 (2): Using general procedure B, cryptophane-233
(0.18 g, 37%) was obtained from compound 20 (0.6 g, 0.52 mmol)
in formic acid (600 mL); decomp. above 300 °C. 1H NMR
(500 MHz, CDCl3, 20 °C): δ ϭ 6.74 (s, 2 H, Ar), 6.68 (s, 2 H, Ar),
6.66 (s, 2 H, Ar), 6.64 (s, 2 H, Ar), 6.59 (s, 2 H, Ar), 6.57 (s, 2 H,
Ar), 4.62 (d, 2J ϭ 14.0 Hz, 6 H, CHa), 4.20Ϫ4.10 (m, 2 H),
4.10Ϫ3.96 (m, 6 H), 3.87 (m, 4 H), 3.82 (s, 6 H, 2 ϫ OCH3), 3.81
(s, 6 H, 2 ϫ OCH3), 3.77 (s, 6 H, 2 ϫ OCH3), 3.41 (d, 2J ϭ 14.0 Hz,
D. J. Cram, M. E. Tanner, C. B. Knobler, J. Am. Chem. Soc.
1991, 113, 7717Ϫ7727.
T. A. Robbins, C. B. Knobler, D. R. Bellew, D. J. Cram, J. Am.
Chem. Soc. 1994, 116, 111Ϫ122.
´
N. Branda, R. M. Grotzfeld, C. Valdes, J. Rebek, Jr., J. Am.
Chem. Soc. 1995, 117, 85Ϫ88.
´
C. Valdes, U. P. Spiz, L. M. Toledo, S. W. Kubik, J. Rebek, Jr.,
J. Am. Chem. Soc. 1995, 117, 12733Ϫ12745.
K. Bartik, M. Luhmer, J.-P. Dutasta, A. Collet, J. Reisse, J.
Am. Chem. Soc. 1998, 120, 784Ϫ791.
T. Brotin, A. Lesage, L. Emsley, A. Collet, J. Am. Chem. Soc.
2000, 122, 1171Ϫ1174.
K. Bartik, M. Luhmer, J. Reisse, A. Collet, J.-P. Dutasta, un-
published results.
S. Mecozzi, J. Rebek, Jr., Chem. Eur. J. 1998, 4, 1016Ϫ1022.
T. Brotin, T. Devic, A. Lesage, L. Emsley, A. Collet, Chem.
Eur. J. 2001, 7, 1561Ϫ1573.
T. Brotin, M. Darzac, D. Forest, M. Becchi, J.-P. Dutasta, J.
Mass Spectrom. 2001, 36, 1092Ϫ1097.
I. Gosse, J.-P. Dutasta, M. Perrin, A. Thozet, New J. Chem.
1999, 23, 545Ϫ548.
J. Canceill, A. Collet, G. Gottarelli, P. Palmieri, J. Am. Chem.
Soc. 1987, 109, 6454Ϫ6464.
2
4 H, CHe), 3.38 (d, J ϭ 13.5 Hz, 2 H, CHe), 2.29 (m, 4 H, CH2)
ppm. 13C NMR (125.67 MHz, CDCl3, 20 °C): δ ϭ 149.22 (2 C),
147.28 (2 C), 147.23 (2 C), 147.20 (2 C), 147.05 (2 C), 146.53 (2
C), 133.71 (2 C), 132.05 (2 C), 131.12 (2 C), 131.05 (2 C), 130.91
(2 C), 130.61 (2 C), 120.26 (2 C), 113.38 (2 C), 112.29 (2 C), 112.19
(2 C), 112.15 (2 C), 112.03 (2 C), 69.42 (2 C), 63.72 (2 C), 63.51 (2
C), 55.82 (2 C, 2 ϫ OCH3), 55.74 (2 C, 2 ϫ OCH3), 55.37 (2 C, 2
ϫ OCH3), 36.26 (2 C, CHa,e), 36.14 (2 C, CHa,e), 35.85 (2 C,
CHa,e), 29.68 (2 C, CH2) ppm. HRMS (LSIMS): exact mass calcd.
for C56H58O12 [M]ϩ 922.3928, found 922.3990.
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
Cryptophane-224 (3): Using general procedure B, cryptophane-224
(0.22 g, 38%) was obtained from compound 22 (0.58 g, 0.51 mmol)
in formic acid (600 mL); decomp. above 300 °C. 1H NMR
(500 MHz, CDCl3, 20 °C): δ ϭ 6.81 (s, 2 H, Ar), 6.72 (s, 2 H, Ar),
J. Canceill, L. Lacombe, A. Collet, J. Am. Chem. Soc. 1985,
107, 6993Ϫ6996.
2
6.69 (s, 2 H, Ar), 6.65 (s, 2 H, Ar), 6.63 (s, 2 H, Ar), 4.61 (d, J ϭ
´
A. Tambute, J. Canceill, A. Collet, Bull. Chem. Soc. Jpn. 1989,
2
14.0 Hz, 4 H, CHa), 4.56 (d, J ϭ 14.0 Hz, 2 H, CHa), 4.38Ϫ4.30
62, 1390Ϫ1392.
(m, 2 H, OCH2), 4.25Ϫ4.18 (m, 2 H, OCH2), 4.15Ϫ4.10 (m, 2 H,
OCH2), 4.07Ϫ3.93 (m, 4 H, OCH2), 3.91Ϫ3.83 (m, 2 H, OCH2),
3.83 (s, 6 H, 2 ϫ OCH3), 3.75 (s, 6 H, 2 ϫ OCH3), 3.70 (s, 6 H, 2
S. Akabori, M. Miura, M. Takeda, S. Yuzawa, Y. Habata, T.
Ishii, Supramol. Chem. 1996, 7, 187Ϫ193.
The rapid loss of efficiency of the Chiralpak OTϩ column did
not allow us to perform a correct analysis of compound 3.
J. Ripmeester, J. Am. Chem. Soc. 1982, 104, 289Ϫ290.
Kindly provided by Prof. Rudkevich and co-workers. Crypt-
ophanes were first minimized with Macromodel 7.1, Amber
force field, then the program GRASP was used to estimate the
volume of the internal cavities of cryptophanes as described
2
ϫ OCH3), 3.40 (d, J ϭ 14.0 Hz, 6 H, CHe), 1.81 (m, 4 H, CH2)
[17]
[18]
ppm. 13C NMR (125.67 MHz, CDCl3, 20 °C): δ ϭ 149.42 (2 C),
148.88 (2 C), 148.55 (2 C), 146.63 (2 C), 146.41 (2 C), 146.25 (2
C), 134.24 (2 C), 133.53 (2 C), 132.16 (4 C), 131.54 (2 C), 131.08
(2 C), 120.16 (2 C), 119.00 (2 C), 117.05 (2 C), 114.80 (2 C), 113.02
(2 C), 112.79 (2 C), 69.65 (2 C), 68.62 (2 C), 68.07 (2 C), 56.71 (2
C, 2 ϫ OCH3), 55.39 (2 C, 2 ϫ OCH3), 55.14 (2 C, 2 ϫ OCH3),
36.19 (2 C, CHa,e), 36.11 (4 C, CHa,e), 24.76 (2 C, 2 ϫ CH2) ppm.
HRMS (LSIMS): exact mass calcd. for C56H58O12 [M]ϩ 922.3928,
found 922.3926.
in ref.[8]
.
[19]
M. M. Spence, S. M. Rubin, I. E. Dimitrov, E. J. Ruiz, D. E.
Wemmer, A. Pines, S. Q. Yao, F. Tian, P. G. Shultz, Proc. Natl.
Acad. Sci. USA 2001, 98, 10654Ϫ10657.
M. Pons, O. Millet, Prog. Nucl. Magn. Reson. Spectrosc. 2001,
38, 267Ϫ324.
[20]
[21]
The matrix L may be calculated numerically by solving the
following system: R ϭ (lnAij)/τm, where Aij ϭ Iij(τm)/M0 is the
j
experimental matrix that contains all the intensities for each
site and for the three selective irradiations.
M. Darzac, T. Brotin, D. Bouchu, J.-P. Dutasta, Chem. Com-
mun. 2002, 48Ϫ49.
Acknowledgments
We thank Prof. D. M. Rudkevich and co-workers for providing us
with the calculations of the inner space volumes of the crypto-
phanes.
[22]
Received September 6, 2002
[O02500]
984
Eur. J. Org. Chem. 2003, 973Ϫ984