Cyclochiral conformers of resorcin[4]arenes stabilized by hydrogen bonds

Article abstract of DOI:10.1039/b701451a

Cyclochiral resorcinarenes, that maintain their cyclochirality by means of hydrogen bonds, were synthesized by a sequence of reactions involving the Mannich reaction, removal of the N,O-acetal bridge and subsequent N-substitution with an RCO group. During

Full text of DOI:10.1039/b701451a

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Cyclochiral conformers of resorcin[4]arenes stabilized by hydrogen bonds†
Agnieszka Szumna*
Received 30th January 2007, Accepted 7th March 2007
First published as an Advance Article on the web 23rd March 2007
DOI: 10.1039/b701451a
Cyclochiral resorcinarenes, that maintain their cyclochirality by means of hydrogen bonds, were
synthesized by a sequence of reactions involving the Mannich reaction, removal of the N,O-acetal
bridge and subsequent N-substitution with an RCO group. During this study it was found that ethyl
nitroacetate is a mild and very efficient agent for N,O-acetal bridge removal. The resulting
resorcinarenes 4a–j exist in cyclochiral/inherently chiral kite conformations (resembling 4-bladed
propellers) that are stabilized by eight hydrogen bonds (in both solid state and solution). It is shown
that the cycloisomerization process is characterized by the relatively high racemization barrier
(14.6–18.5 kcal mol−1 as determined by 2D EXSY) and thus it can be concluded that the
transformation of one cycloconformer into the other requires the simultaneous rupture of all eight
hydrogen bonds. For derivatives with additional stereogenic centers two cyclodiastereoisomeric
conformations were detected (diastereomeric excess in the range of 72% up to >95%). The experimental
results are additionally supported by AM1 semi-empirical calculations.
compounds made of the same building blocks that differ only
Introduction
in the sense of direction of the ring.7,8 Although, according
For the construction of chiral ligands for a variety of applications
to strict definitions, these two terms do not overlap, in the
(including catalysis, separation, recognition and detection) the use
literature for some types of compounds the terms have been
of stereogenic centers as sources of chirality is a classical and still
used interchangeably. For the purpose of this paper the term
dominant approach. However, recently other types of chirality like
cyclochirality will be used as the one that is more in keeping with
axial1,2 or planar3 have also gained increasing importance. Not
long ago the first application of molecules possessing so called
‘inherent chirality’ or ‘cyclochirality’ was reported.4,5 These types
the phenomenon that is discussed.
The calix[n]arene skeleton is well suited for the construction
of cyclochiral/inherently chiral molecules due to its immanent
of chirality are dependent upon the curvature of the molecules.
curvature that can be preserved. Several classes of calixarenes
Inherent chirality derives from curvature of molecules that are
possessing these types of chirality have been reported. They involve
devoid of symmetry axes in their bi-dimensional representation.6
The other similar term—cyclochirality—has been defined for
calix[4]arenes with either a WXYZ or XXYZ substitution pattern
at the upper rim, I,9 or cyclochirally substituted resorcin[4]arenes,
II and III (Chart 1).10,11 The attractiveness of the last two types of
Institute of Organic Chemistry, Polish Academy of Sciences, Kasprzaka
44/52, 01-224, Warsaw, Poland. E-mail: szumnaa@icho.edu.pl; Fax: +48
22 632 6681; Tel: +48 22 343 2110
† Electronic supplementary information (ESI) available: crystallographic
data for 4j in CIF format, NMR spectra (PDF) and AM1 optimized
structures in ML2 format. See DOI: 10.1039/b701451a
compounds lies in the relative easy synthesis (using the Mannich
reaction) and resolution as compared with substituted calixarenes
I. The cyclochiral rim of intramolecular hydrogen bonds in
resorcinarenes II and III drives the Mannich reaction to one
regioisomer only and in the case of using chiral amine often to a
Chart 1 Examples of cyclochiral/inherently chiral calix[4]arenes.
1358 | Org. Biomol. Chem., 2007, 5, 1358–1368 This journal is
The Royal Society of Chemistry 2007
©

View Article Online
single diastereoisomer.12 Unfortunately, simple Mannich products
II suffer from instability due to facile epimerization at the N,O-
acetal bridge.13 The stability problem has been overcome for
compounds of type III.
The current paper deals with cyclochiral compounds in which
cyclochirality is of non-covalent origin. The synthesis, structure
and stability of conformationally cyclochiral resorcinarenes 4
(Scheme 1) that maintain their cyclochirality by means of hydrogen
bonds will be discussed. Additionally it will be shown that even
though the cyclochirality is based on non-covalent interactions
considerable diastereomeric excess can be detected.
While the conformational cycloenantiomerism of sterically in-
terlocked molecules has attracted attention for decades, examples
of similar phenomena based on hydrogen bonds are limited. Early
examples were reported by the group of Rebek.14–16 They created
deep cavitands with “doors” at the upper rim that were controlled
by a directional cooperative belt of hydrogen bonds, formed
between the amide or hydroxy groups. For cyclodiastereomeric
compounds diastereomeric excess was observed (d.e. 50%).
Another example involves an upper rim octaurea-functionalized
resorcinarene, for which the existence of the cyclochiral belt
of eight hydrogen bonds has been postulated.17 Calix[4]arene
substituted at the lower rim by amino acids was also postulated to
form a cyclochiral cooperative belt of four hydrogen bonds.18
Another interesting example, reported by Schmidt et al.,19 is
actually a prototype of the system studied in this paper. Authors
reported on the accidental synthesis of the N-acetyl derivative of
the Mannich product obtained from an amine and tetraethylre-
sorcin[4]arene. It was noticed that two distinct hydroxyl groups
were observed by NMR which was indicative of cyclochiral
conformation (as will be discussed later). However, no detailed
conformational or stability studies were performed for this system.
Scheme 1 Synthesis of conformationally cyclochiral resorcinarenes: (a) amine, HCHOaq, AcOH, MeOH; (b) ethyl nitroacetate, MeOH, rt, 2 h; (c),(d),(e)
see Table 1 and Experimental section.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1358–1368 | 1359
©

View Article Online
The tert-butylaminocarbonyl substituent was introduced using
tert-butyl isocyanate. The reaction required a coarse purification
of the crude reaction mixture containing 3a–e by washing with
hexane and vacuum drying. The resulting oil was dissolved in
THF and treated with tert-butyl isocyanate. Products 4f–j were
isolated by column chromatography.
The direct pathway, without removal of the aminoacetal bridge
(Scheme 1, pathway e), was also studied. Two procedures were
examined: direct reaction in THF and a two-phase reaction
CH2Cl2/water. While for acetyl derivatives under two-phase
conditions (CH2Cl2/water, Table 1, entry 3) route e gave the desired
product 4a in reasonable yield, for other derivatives (Boc and Bac)
no products were observed. The other direct pathway (Scheme 1,
route d), i.e. the Mannich reaction of N-Boc protected amines and
formaldehyde with resorcinarenes 1a,b, was also pursued. It did
not give products either.
All products 4a–j obtained in such a way have selectively
substituted four N atoms, while the remaining hydroxy groups
remain unsubstituted. For compounds that have eight OH and
four amide groups the products are surprisingly non-polar. For
example they are soluble in hexane, while they usually precipitate
from MeOH, MeCN or even DMF and DMSO. This allows
the assumption that for all products polar groups are engaged
in intramolecular hydrogen bonding interactions, most probably
following the pattern that was reported by Bo¨hmer.19
Results and discussion
Synthesis
In the current study I have synthesized resorcinarenes with acetyl
(Ac), tert-butoxycarbonyl (Boc) and tert-butylaminocarbonyl
(Bac) substituents on the nitrogen atom (Scheme 1). Tetralkyl
resorcinarenes 1a–b were transformed into tetrabenzoxazines 2a–
d by the Mannich reaction of 1a–b with formaldehyde and ben-
zylamine or (R)-1-phelylethylamine in methanol using a known
procedure.10,12 The desired products 2a–d precipitated from the
reaction mixture and were isolated in high yields. While in the work
by Bo¨hmer,19 the N-acetylation of tetrabenzoxazines was achieved
by chance and in moderate yield (18%), I have sought an efficient
and general method for the N-substitution of tetrabenzoxazines
2a–d. It seemed reasonable to cleave the N,O-acetal bridge first and
then perform the N-substitution reaction. The known method for
N,O-acetal removal involves using HCl in butanol (reflux). Such
a method is efficient but it cannot be considered mild. During the
course of this work it was found that ethyl nitroacetate reacts with
tetrabenzoxazines 2a–d with removal of the N,O-acetal bridge
(the reaction was completed after ca. 2 h in methanol at room
temperature). Additionally, it produces free amines 3a–d and thus
does not require using a base at the next stage (which could have
caused undesired O-substitution). In all cases careful isolation and
purification of free amines 3a–d from the reaction mixture was not
required. Thus the yields of 4a–j given in Table 1 are for the two
steps b and c.
X-Ray structure
The tert-butoxycarbonyl substituent was introduced using two
different procedures. The convenient general procedure involves
addition of Boc2O to the crude reaction mixture containing amines
3a–e in methanol affording precipitates, which upon recrystalliza-
tion from CH2Cl2–MeCN gave analytically pure compounds 4a–e
in high yield (e.g. in the case of 4b in 99% yield, Table 1, entry
5). The reaction could be also carried out “on water” (Table 1
entry 4). In this case, the reaction mixture containing 3b had to
be evaporated prior the addition of water and Boc2O. Such a
green approach was earlier reported to produce selectivity of the
amino over phenol/alcohol Boc protection of amino acids and
other simple amines.20 Using this approach the desired product
4b was obtained in 85% yield (Table 1, entry 4). Although this
method is “green” it is far less convenient at the lab scale due to
the requirement of hand-grinding of the reaction mixture.
The X-ray structure of compound 4j confirmed an extensive
intramolecular hydrogen-bonding pattern (Fig. 1). The molecule
of 4j has C4 symmetry (both molecular and crystallographic) and
adopts a cyclochiral kite conformation that is stabilized by the
presence of eight hydrogen bonds. The most characteristic feature
of this structure is the presence of four eight-membered hydrogen
bonded rings that can bond concurrently in only one direction and
thus are responsible for the existence of cyclochiral conformation.
Within the eight-membered ring the strong hydrogen bonds are
=
formed between the phenolic –OH and the amide C O (distances:
˚
˚
O8 · · · O13 2.58 A, H8 · · · O13 1.78 A, angle O8–H8 · · · O13
154◦). Outside the ring the same phenolic –OH group act as a
hydrogen bond acceptor for the next slightly longer interaction
˚
with the neighbouring phenolic –OH (distances: O9 · · · O8 2.70 A,
◦
˚
H9 · · · O8 2.07 A, angle O9–H9 · · · O8 132 ).
Table 1 Synthesis of cyclochiral resorcin[4]arenes 4 (Scheme 1)
Entry
Product
R1
R2
R3
(Steps) Conditions
Yield (%)
1
2
3
4
5
6
8
9
10
11
12
13
14
4a
i-Bu
Bn
methyl
(c) MeOH
(c) THF
(e) water/CH2Cl2
(c) water
(c) MeOH
(c) MeOH
(c) MeOH
(c) MeOH
(c) THF
(c) THF
(c) THF
(c) THF
(c) THF
56
21
43
85
99
61
56
56
63
25
54
48
30
4b
i-Bu
Bn
tert-butoxy
4c
4d
4e
4f
4g
4h
4i
i-Bu
i-Bu
Et
i-Bu
Et
i-Bu
i-Bu
Et
n-Bu
tert-butoxy
(R)-1-phenylethyl
(R)-1-phenylethyl
Bn
tert-butoxy
tert-butoxy
tert-butylamino
tert-butylamino
tert-butylamino
tert-butylamino
tert-butylamino
Bn
n-Bu
(R)-1-phenylethyl
(R)-1-phenylethyl
4j
1360 | Org. Biomol. Chem., 2007, 5, 1358–1368
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
environment, which manifests in its small temperature coefficient
(Dd/DT 4.7 × 10−4 ppm K−1) in NMR experiments.
It could be envisioned that the epimerization of such a system
(to give a second cyclodiastereoconformer) would be difficult,
since the inversion at one ring only would have to disturb the
entire belt of hydrogen bonded atoms and cause unfavourable
interactions. This should cause a considerable kinetic stability for
the conformers (i.e. high barriers to cycloepimerization).
NMR studies
1
Room temperature studies. For all products 4a–j H and 13C
NMR spectra in CDCl3 and toluene-d8 exhibit a reduced number
of signals in accordance with the C4 symmetry of the molecules. In
all cases two different signals for –OH protons were observed with
a large chemical shift difference (i.e. one at ca. 8.9 ppm and the
second one in the range of 11.5–12.8 ppm). The existence of two
distinct –OH groups indicates that the molecules exist in relatively
stable (exchange is slow on the NMR timescale) cyclochiral
conformations with overall C4 symmetry not only in the solid
state but also in solution. The chemical shifts of those protons are
considerably downfield shifted and concentration independent.
This suggests that the intramolecular hydrogen bonding pattern
observed in the solid state is most probably also maintained in
solution.
Compounds 4a–c and 4f–h, devoid of other types of chirality,
exist in the solution as mixtures of (P)- and (M)-cycloenantiomeric
conformations (see Scheme 1). However, for compounds with ad-
ditional stereocenters 4d–e and 4i–j, two cyclodiastereoconformers
(P)-(R)- and (M)-(R)-can be expected having most probably
different spectra and unequal populations.
In fact for (R)-1-phenylethylamine-Bac derivatives 4i and 4j two
1
sets of signals can be extracted from H and 13C NMR spectra
in toluene-d8 and in CDCl3 with a ratio of 0.86 : 0.14 (d.e.
72%). This transforms into 1.1 kcal mol−1 difference in stability
of the cycloconformers. The NOESY and ROESY spectrum of
4i (see ESI†) shows the presence of cross-peaks between two
diastereotopic methylene protons and the methine proton at the
stereogenic centre: Hf1 ↔ Hg and Hf1 ↔ Hg (Fig. 2, see also
ESI†). This may indicate that this proton is directed toward the
upper side of the molecule (the opposite to what was observed
in the solid state). The spectrum also shows the weak cross-
peaks between the amide NH and all protons of the neighbouring
Fig. 1 The X-ray structure of (M)-(R)-4j: a) top view; b) side view. An
acetonitrile (66%) or methanol (34%, not shown) molecule occupies the
cavity.
Since the molecule contains additional known stereogenic
centers ((R)-1-phenylethylamine fragments), the cyclochirality of
4j can be determined unambiguously by the direction of the eight-
membered hydrogen bonded ring closure. Assuming that: (a) the
lower rim of the resorcin[4]arene 4j is the bottom of the molecule,
(b) looking from within the cavity, and (c) assuming priority of the
–OH group hydrogen-bonded to the carbonyl oxygen atom over
the –OH group hydrogen-bonded to the phenolic oxygen atom, the
direction of ring closures can be described as anticlockwise or M
(denoted as (M)-(R)-4j). It should be noted at this point that there
is no uniform notation for the stereochemistry of cycloisomers
of that type. Although it seems that an approach of looking from
within the cavity is more spread out, an alternative view by looking
from above the resorcinarene ring has recently been proposed but
gives the opposite result.21 In this paper the convention of looking
from within the cavity is used uniformly.
The other important feature of this structure is that the amide
hydrogen atom is in close proximity to the aromatic ring (as was
also deduced from NMR spectra, see next section). However the
amide proton is not forming a N–H · · · p interaction but instead
points into a void formed by the methyl group and the phenyl
ring. Despite this, the amide N–H remains quite shielded from the
Fig. 2 Relevant ROESY contacts for major cycloconformer of 4i
(cycloconformation is ascribed arbitrarily).
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1358–1368 | 1361
©

View Article Online
(R)-1-phenylethylamine part: NH ↔ Hg, NH ↔ Hh, NH ↔
Hi (Fig. 2, see also ESI†). This is due to the flexibility of
(R)-phenylethylamine and unfortunately prevents unambiguous
determination of the relative configuration of the cycloconformer.
For tert-butoxycarbonyl compounds 4d and 4e, possessing also
central chirality, only one set of signals in the spectra can be found
which implies that either: 1) one diastereoconformer is formed
predominately (d.e. >95%); or 2) the exchange is dramatically
faster than for other compounds; or 3) signals accidentally overlap.
The dramatically faster exchange seems to be unlikely since two
OH signals are still visible. An accidental overlap of signals cannot
be excluded unambiguously, however it should be noted that for
4e signals are relatively sharp (even sharper than for 4j and scalar
splitting, e.g. for AB systems, can be seen clearly). Additionally, the
13C NMR spectrum shows one set of signals (see ESI†) and thus
the accidental overlap of signals also appears to be unlikely. Thus,
for 4e it can be postulated that a single cyclodiastereoconformer
exists predominately (d.e. >95%) in CDCl3 solution at 303 K. The
NOESY spectrum of 4e shows the presence of weak cross-peaks
between diastereotopic methylene Hf ↔ Hi (aromatic), Hf ↔
Hh (methyl protons) and Hf ↔ Hg (methine) (Fig. 3). This is
due to partially free rotation of the (R)-phenylethylamine group
and unfortunately prevents the determination of the relative
configuration of the cycloconformer.
thus would result in signal broadening and finally coalescence.
The shape analysis of the variable temperature NMR spectra
(VT NMR) should thus give the rates of exchange. However,
in this case the VT NMR spectra (303–373 K interval) show
that increasing the temperature causes the half-widths of the
OH signals to first slightly decrease and then increase (Fig. 4).
Additionally it was observed that at room temperature the half-
widths of the OH signals are strongly magnetic field dependent.§
Such behavior can be explained if one assumes that at room
temperature each cycloconformer is in fast equilibrium with a
set of its low-abundance conformers which derive from breaking
of individual hydrogen bonds (one or more), however, without
significant structure rearrangement. At room temperature the
exchange by “individual hydrogen bond rupture” is relatively
faster than the exchange by “racemization” and thus the former
contributes mainly to the line broadening. As the temperature
increases, the exchange by “individual hydrogen bond rupture”
becomes very fast, while the exchange by “racemization” reaches
the region of intermediate rates and thus is the main contributor
to the line broadening.
Fig. 4 Temperature dependence of half-width of OH signals (average of
two OH signals).
Fig. 3 Relevant NOESY contacts for 4e (cycloconformation is ascribed
arbitrarily).
Due to the additional co-existing process (being in the regime
of fast exchange) it is difficult to extract kinetic data from
these measurements. However some qualitative conclusions can
still be elucidated, especially from the high temperature data
(Fig. 4). From the data, it is worth noting that compounds 4b–
d having tert-butoxycarbonyl substituents reach coalescence at ca.
365 K, while tert-butylaminocarbonyl derivatives 4f and 4i do not
reach coalescence up to 373 K. Additionally the half-widths of
signals for tert-butylaminocarbonyl derivatives (which is directly
proportional to the exchange rate) at a given temperature (e.g.
353 K) are considerably lower than for their tert-butoxycarbonyl
counterparts (4b vs. 4f and 4d vs. 4i). This means that compounds
having NH atoms instead of an oxygen atom have more stable
cycloconformations. This may be attributed to the bulkiness of
the NH group compared with the oxygen atom and/or favourable
interactions of NH with the neighbouring aromatic ring as
compared with the probably repulsive interaction between an
Kinetic studies. The process of racemization between two
cycloconformers involves an exchange between distinct hydroxy
protons. The rates of exchange transform directly into racem-
ization barriers, which can be used as measures of the relative
tendency of molecules to maintain their intramolecular hydrogen
bonding and thus their cyclochiral conformation. The exchange
process was studied using variable temperature measurements and
2D EXSY.
The variable temperature measurements were performed in
toluene-d8 for 4b–d, 4f and 4i. To minimize experimental errors
the experiment was conducted so that the whole set of samples
was recorded at the same time at a given temperature after
stabilization.‡ For signals of protons in a slow to intermediate
exchange regime (on the NMR timescale) it is expected that
increasing the temperature would speed up the exchange and
‡ It has been checked experimentally that this method produces less
experimental errors than measuring each sample separately.
§ For example for 4b the half-width of OH signals are 10 Hz on 200 MHz
NMR and 40 Hz on 400 MHz NMR at 295 K.
1362 | Org. Biomol. Chem., 2007, 5, 1358–1368
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
oxygen atom and an aromatic ring. The proximity of NH atoms
to the benzylamine aromatic ring is also supported by the X-ray
structure and by the chemical shift of the NH protons, which
show up at 3.94 ppm for 4i (assignment based on 1H–15N HSQC
spectrum). Such an upfield shift is most probably due to shielding
caused by the neighboring aromatic ring. Steric hindrance at
the benzylamine part also has a pronounced, though smaller,
positive effect on the stability (i.e. compare derivatives of (R)-
1-phenylethylamine 4d and 4i with their respective benzylamine
analogues 4b and 4f).
minimize errors caused by co-existing exchange) the extracted rate
constants from variable temperature spectra and extrapolated to
303 K give k = 1.0 s−1 in toluene-d8 which is in surprisingly good
agreement, with those obtained from the EXSY spectrum. The
rate constant k = k1 + k−1 = 1.9 s−1 for an 0.86 : 0.14 conformer
ratio translates to exchange barriers of DG2 = 18.5 kcal mol−1
(major → minor) and 17.4 kcal mol−1 (minor → major) at
303 K in CDCl3. Those are quite reasonable values for the cost
of rupturing simultaneously eight hydrogen bonds. The typical
costs of hydrogen bond rupture in organic solvents is roughly 1 to
2 kcal mol−1 per hydrogen bond.23 In fact for the breaking of eight
coherent hydrogen bonds in cavitands a barrier of 17.4 kcal mol−1
was observed. However, it should be noted that in the previous
case the hydrogen bonds are arranged in a cooperative seam (i.e.
one amide group that is a donor of hydrogen bond simultaneously
acts as an acceptor for the next bond). In contrast this system has
a hydrogen bond system that is spatially separated and divided
into four parts, which are interlocked by rotation. It is also
important to note that the existence of a cooperative seam of eight
hydrogen bonds is not enough for the considerable kinetic stability
of cycloconformers (e.g. octaurea-functionalized resorcinarenes
reveal rapid interconversion between conformers).
The 2D EXSY spectrum of 4e ((R)-phenylethylamine-Boc
derivative) shows the exchange cross-peaks only between the
two different OH groups (in two spectra that used different
mixing times sm = 20 ms and sm = 250 ms, Fig. 6). The exchange
process that leads to entities that are indistinguishable from the
starting molecules has the disadvantage that the exchange intensity
data can be strongly biased by NOE (OH protons are in close
proximity). To avoid this the kinetic data were extracted from the
EXSY spectrum using a short mixing time sm = 20 ms. A short
mixing time causes the NOEs to be negligible and fortunately the
intensities of the exchange cross-peaks were reasonably high. How-
ever, the inability to observe the second diastereoconformer raises
the question about interpretation of kinetic data, since the con-
version to the identical cyclodiastereoconformer, observed in the
NMR experiment, requires two rotation steps (i.e. (M)-(R)-4e →
(P)-(R)-4e → (M)-(R)-4e). Additionally calculation of rate con-
stants from the EXSY spectrum requires knowledge of the mole
fractions, so the assumption of the ratio of diastereoconformers
The exchange process for 4i ((R)-phenylethylamine-Bac deriva-
tive) was observed using 2D EXSY spectroscopy. In this case
racemization of the cyclic array of hydrogen bond produces
a second diastereoconformer, and thus the exchange process
should take place between the two cyclodiastereoconformers. The
existence of separate signals for two diastereoconformers is a
lucky situation since it allows for elimination of NOE effects,
which are otherwise difficult to separate. The exchange correlation
cross-peaks appear between signals of different cycloconformers
(Fig. 5). The basis for the extraction of kinetic data is integration
of the cross-peak volumes in the 2D EXSY NMR spectra.22 The
exchangerate, calculatedusing integration ofsignals from different
protons undergoing exchange in 4i (OH, NH and aliphatic), was
found to be k = 1.9
0.3 s−1 at 303 K in CDCl3. It should
be noted that at optimal temperatures (high temperature range to
Fig. 5 2D EXSY spectrum of 4i: a, major cycloconformer; b, minor
cycloconformer (CDCl3, 500 MHz, 303 K, sm = 250 ms). a) Amide NH
region; b) OH region.
Fig. 6 2D EXSY spectrum of 4e (CDCl3, 500 MHz, 303 K, sm = 20 ms);
OH protons region.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1358–1368 | 1363
©

View Article Online
has to be made prior to the calculations. Since the exchange is
visible, the second diastereconformer has to exist in a detectable
amount. However, its signals would have: (1) lower intensity and
(2) larger half-width due to faster exchange and thus might not be
visible. Based on the condition d.e. ∼95% and on the calculation
mathematical limits the assumption of a 0.976 : 0.026 conformer
ratio was made. The exchange rate was calculated to be k = k1 +
k−1 = 195.2 s−1 at 303 K, which transforms into interconversion
barriers of 16.7 kcal mol−1 (major → minor) and 14.6 kcal mol−1
(minor → major) at 303 K in CDCl3.
CD spectra
Fig. 8 Preset for AM1 optimizations.
Circular dichroism spectroscopy provides further evidence for
the existence of stable cyclodiastereoconformers. Fig. 7 show the
CD spectra of 4d, 4e, 4i, and 4j. The existence of pronounced,
though not very strong (as compared with other “covalently”
cyclochiral resorcinarenes)24 Cotton effects of opposite signs
was observed for the absorption bands at 291 and 310 nm,
which are attributed to electronic transitions of the resorcinol
chromophore. It indicates that even though the stereogenic centers
are quite remote, the parent resorcinarene ring still experiences
a chiral environment. This is probably due to the cyclochiral
arrangement of hydrogen bonds. In fact, Cotton effects caused
by cyclochiral conformation have been reported previously for
ocataurea-substituted resorcinarenes.
by 30◦ and the resulting structures were optimized. The KITE
conformations are obtained when the eight-membered ring points
its N atom out from the molecular cavity (T3 = −90◦), while
VASE conformations were generated by pre-setting the N atom
to be directed inward (T3 = 90◦). During the optimization the
torsion angles T1 underwent changes but a change from (M) to
(P) configuration or vice versa was never observed.
Table 2 presents the results of the AM1 optimizations. Most
optimized conformations have C2 symmetry and the slightly
flattened-cone conformation of the parent resorcin[4]arene ring.
Although NMR and X-ray analysis shows the C4 symmetry,
the result of the calculations is not in contradiction since the
interconversions between two C2 conformations are probably
facile, thus the time-averaged conformation has a C4 symmetry.
For the least crowded acetyl derivative 4k the AM1 calculations
indicate that the VASE-(M) is most stable conformation. It should
be noted that the number of low energy stable conformations
is relatively larger than for 4l and 4m. This indicates that the
flexibility of this system is considerably larger than for other
derivatives.
For 4l ((R)-phenylethylamine-Boc, an analogue of 4d and
4e), AM1 calculations gave predominately two conformations
KITE-(P) and VASE-(M) (Table 2). The most stable KITE-(P)
conformation is the only one with low energy among the KITE
conformations. In fact NMR spectra for analogous compound
4e suggested the presence of a single diastereoconformer. The
stability of the VASE-(M) conformation is rather surprising since
it is tightly packed with all four methyl group closing the upper rim
of the molecule. To double-check the result an extensive additional
search for low energy KITE-4l conformations was performed
(other dihedral angles were also varied and additionally non-
symmetric conformations were also checked) but did not yield any
conformation that was different from the one that was originally
found.
Fig. 7 CD spectra of cyclochiral resorcinarenes (CHCl3, 295 K).
Based on the similarity of the CD spectra (Fig. 7), it can
be speculated that in all cases the same cycloconformer is
predominant. The larger values of De for tert-butoxycarbonyl
derivatives as compared with tert-butylaminocarbonyl derivatives
may also indicate that in the former case, the diastereomeric
excess of a particular diastereoconformer is higher in CHCl3
solution. Although this reasoning is speculative, it is in reasonable
agreement with NMR data (see previous section).
For molecule 4m ((R)-phenylethylamine-Bac, an analogue of
compounds 4i and 4j), AM1 calculations gave two stable KITE
cycloconformers: KITE-(M)-(R)-4m (68%) and KITE-(P)-(R)-
4m (25%). Thus the calculations predicted higher stability of
the M-cycloconformer i.e. the same as observed in the solid
state for analogous compound 4j. The diastereoconformer ratio
is also in rough agreement with the NMR data (for 4i two
diastereoconformers were observed with ratio 86 : 14). However, it
should be pointed out that in the case of the molecule having tert-
butylaminocarbonyl substituents the most stable cycloenantiomer
istheoppositetotheoneforthemoleculewith tert-butoxycarbonyl
groups.
AM1 semiempirical calculations
The semiempirical calculations (AM1) were performed for deriva-
tives of (R)-1-phenylethylamine (Fig. 8). The only simplification
compared to real compounds consisted of using methyl, instead
of iso-butyl or ethyl group at the lower rim of resorcin[4]arenes,
which is believed to have a minor impact on the conformational
preferences. In order to screen a conformational space for each
compound KITE-(M), KITE-(P), VASE-(M) and VASE-(P) were
preset (Fig. 8) and then for each conformation the torsion angles
between the amide part and the benzylamine part (T1) were varied
1364 | Org. Biomol. Chem., 2007, 5, 1358–1368
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
Table 2 Semiempirical AM1 optimization results
◦
Comp.
Conformation
T1out
/
Relative energy/kcal mol−1
Distribution at 295 K
4k
VASE-(M)
KITE-(P)
VASE-(M)
VASE-(M)
VASE-(M)
KITE-(M)
KITE-(M)
VASE-(P)
KITE-(M)
VASE-(P)
KITE-(M)
VASE-(P)
VASE-(P)
KITE-(P)
−65, −75 (×3)
0
0.4486
0.1816
0.1248
0.0725
0.0681
0.0626
0.0209
0.0188
0.0013
0.0002
<0.0001
<0.0001
<0.0001
<0.0001
−68
0.53
0.75
1.07
1.11
1.15
1.77
1.86
3.42
4.29
5.07
5.19
5.82
6.57
−75
−93, −75 (×3)
−60
−60
−51
96
96
74
71
98, −52, 72, −52
−54
74
4l
KITE-(P)
VASE-(M)
KITE-(M)
VASE-(P)
KITE-(M)
KITE-(P)
−81
−81
88
0
0.5093
0.4855
0.0007
0.0043
<0.0001
<0.0001
0.03
3.85
2.79
5.52
9.84
86
−71
92
4m
KITE-(M)
KITE-(P)
VASE-(M)
VASE-(P)
VASE-(P)
−60
−66
−64
−55
66
0
0.6821
0.2484
0.0695
<0.0001
<0.0001
0.59
1.34
5.71
9.23
It should be emphasized that the results of these calculations
should be interpreted with great caution. Although it seem
intuitive that the KITE conformations should more stable due
to the lack of steric overcrowding, the calculations suggest that
the VASE conformations are not unthinkable, and even they
can be quite stable. However, the internal volumes of the VASE
conformations are very small and usually their “doors” at the
upper rims are tightly closed.
intermediates and thus for racemization simultaneous rotation of
all units is required.
Although these systems are not very promising for complexation
due to their KITE conformations resulting in relatively shallow
cavities, they offer an interesting insight into a restricted molecular
motion of the propeller-like chiral molecules, which are of
particular interest as molecular rotors.25
Experimental
Conclusions
All chemicals were used as received unless otherwise noted.
Reagent grade solvents (CH2Cl2, hexanes) were distilled prior to
use. All reported H NMR spectra were collected on a Bruker
spectrometer at 500 (1H NMR) and 125 (13C NMR) MHz.
Chemical shifts are reported as d values relative to the TMS signal
defined at d = 0.00 (1H NMR) or relative to the CDCl3 signal
defined at d = 77.00 (13C NMR). Mass spectra were obtained
on a Mariner PerSeptive Biosystem instrument using the ESI
technique. Chromatography was performed on silica gel (Kieselgel
60, 200–400 mesh).
The resorcinarenes 1a,b can be conveniently transformed into
upper rim amido substituted resorcin[4]arenas 4a–j using a
sequence of reactions: the Mannich reaction, removal of the
N,O-acetal bridge and subsequent N-substitution with an RCO
group. The newly discovered efficient procedure of N,O-acetal
removal using ethyl nitroacetate is very mild as compared with the
known procedure using refluxing in HCl and butanol. This may
be advantageous for the synthesis of more sensitive compounds
(e.g. hydrolysis-prone amino acid derivatives).
1
Amide substituted resorcin[4]arenes 4a–j exist as conforma-
tional cycloisomers stabilized by belts of hydrogen bonds. The ex-
change between cycloconformers is slow on the NMR timescale. It
allows for determination of the racemization barriers and relative
stability of the cyclodiastereoconformers. Based on relatively high
racemization barriers it can be concluded that the cycloisomeriza-
tion process proceeds through simultaneous rupture of all eight
hydrogen bonds. This is a quite surprising result assuming that the
hydrogen bond system is spatially divided into four parts, i.e. each
part consists of just two hydrogen bonds. However, rotation of just
one unit creates an unfavourable pattern of the adjacent hydrogen
bonds, which is enough to prevent stability (and detection) of such
The semiempirical calculations were preformed using Gaussian
03 software.26
CD spectroscopy
CD spectra were measured at room temperature in chloroform (for
UV spectroscopy, Fluka) with solutions with concentrations of 1 ×
10−4 M on a Jasco 715 spectrophotometer by using cells with path
length 0.1 to 1 cm (spectral band width 2 nm, sensitivity 5 × 10−6
or 10 × 10−6 (DA-unit nm−1), where DA = AL − AR is the difference
in the absorbance). De is expressed in units of L mol−1 cm−1.
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1358–1368 | 1365
©

View Article Online
NMR experiments
4 H, OH). dC (125 MHz, CDCl3, 303 K): 22.85, 22.96, 26.68, 31.75,
42.09 (br), 42.80, 53.01, 111.65, 123.67, 124.66, 126.39, 127.46,
128.85, 136.53, 150.25, 151.72, 173.91. ESI MS m/z Calcd for
[C84H100N4O12 − H]− 1355.7; found 1355.7, isotope profile agrees.
Anal. Calcd for C96H128N4O16: C 74.31, H 7.42, N 4.13%; found C
74.15, H 7.32, N 4.30%.
4b. Method A, yield 99%. dH (500 MHz, CDCl3, TMS, 303 K):
0.98 (m, 24 H, Ha), 1.25–1.45 (bs, 36 H, C(CH3)3), 1.51 (m, 4H,
Hb), 2.10 (bs, 8H, Hc), 4.20–4.75 (bm, 20H, Hf, Hd, Hg), 7.20–
7.28 (m, 24H, Har), 8.65 (bs, 4 H, OH), 11.04 (bs, 4 H, OH).
dC (125 MHz, CDCl3, 303K): 22.76, 22.76, 26.11, 27.42, 28.26,
31.80, 40.5 (br), 42.90, 50.83, 81.64, 112.21, 123.56, 124.49, 124.83,
126.94, 127.62, 128.29, 138.47, 150.29, 151.63, 158.78. ESI MS
m/z Calcd for [C96H128N4O16Na]+ 1611.9; found 1612.0, isotope
profile agrees. Anal. Calcd for C96H128N4O16: C 72.33, H 8.09, N
3.51%; found C 72.23, H 7.92, N 3.68%.
Variable temperature NMR spectra were recorded on a 500 MHz
Bruker spectrometer equipped with a BTO2000 thermocouple.
The spectra were recorded in such a way that the whole set of
samples was recorded at a given temperature after stabilization at
the same time (roughly).
The EXSY spectra were recorded at 303 K at 500 MHz on
a Bruker spectrometer with a phase sensitive NOESY pulse
sequence supplied with the Bruker software. The integration of
the 2D spectra was preformed using XWin-NMR software. From
the integrals of the cross-peaks and the diagonal peaks and mole
fractions the rate constants were calculated according to known
equations.22
The general procedure for the Mannich reaction
4c. Method A, yield 61%. dH (500 MHz, CDCl3, TMS, 303 K):
0.91 (t, J = 7.3 Hz, 12 H, CH2-CH3), 0.96 (d, J = 6.5 Hz, 24 H,
Ha), 1.29 (m, 4H, Hb), 1.32–1.52 (m, 52 H, CH2, C(CH3)3), 2.02
(bs, 8H, Hc), 3.32 (bs, 8H, N-CH2), 4.31 (bm, 8 H, Hf1,2), 4.53
(bt, 4 H, Hd), 7.17 (s, 4H, Har), 8.50 (bs, 4 H, OH), 10.93 (bs,
4 H, OH). dC (125 MHz, CDCl3, 303 K): 13.86, 19.98, 22.83,
26.16, 27.47, 28.43, 30.35, 31.85, 40.92, 43.10, 47.36, 80.88, 112.45,
123.51, 124.38, 124.74, 150.26, 151.65, 158.79. ESI MS m/z Calcd
for [C84H132N4O16Na]+ 1476.0, found 1476.0, isotope profile agrees.
4d. Method A, yield 56%. [a]D = +93.3 (c 1.13, CHCl3). dH
(500 MHz, CDCl3, TMS, 303 K): 0.90 (bd, 12 H, Ha), 0.96 (d,
J = 6.6 Hz, 12 H, Ha), 1.52 (m, 4H, Hb), 1.64 (bs, 12H, Hh),
2.11 (bm, 8H, Hc), 4.50–4.56 (m, 12H, Hf, Hd), 5.10, (q, J =
6.8 Hz, 4H, Hg), 7.01 (m, 8H, Hi), 6.98–7.25 (m, 24H, Har), 8.54
(bs, 4 H, OH), 10.96 (bs, 4 H, OH). dC (125 MHz, CDCl3, 303 K):
16.12 (br), 22.67, 22.78, 26.29, 27.42, 28.01, 31.76, 42.78, 43.15
(br), 56.73, 81.41, 112.24, 123.71, 124.52, 124.65, 126.09, 126.37,
127.51, 142.71, 150.11, 151.41, 158.51. ESI MS m/z Calcd for
[C100H132N4O16H]+ 1646.0, found 1646.5, isotope profile agrees.
4e. Method A, yield 56%. [a]D = +140.4 (c 1.19, CHCl3). dH
(500 MHz, CDCl3, TMS, 303 K): 0.94 (t, J = 7.0 Hz, 12 H, Ha),
1.10 (s, 36 H, Hj), 1.52 (m, 4H, Hb), 1.64 (bs, 12H, Hh), 2.27 (m,
8H, Hb), 4.33 (m, 4 H, Hd), 4.49 (AB/2, JAB = 20.9 Hz, 4 H, Hf1),
4.54 (AB/2, JAB = 20.9 Hz, 4H, Hf2), 5.08 (q, J = 6.8, 4 H, Hg),
6.99–7.06 (m, 20H, Hi), 7.26 (s, 4 H, He), 8.57 (bs, 4 H, OH),
10.89 (bs, 4 H, OH). dC (125 MHz, CDCl3, 303 K): 26.84, 28.05,
36.27, 43.12, 56.67, 81.34, 112.21, 123.19, 124.46, 124.54, 126.13,
126.46, 127.59, 142.77, 150.23, 151.60, 158.52. ESI MS m/z Calcd
for [C92H116N4O16Na]+ 1555.8, found 1555.8, isotope profile agrees.
Anal. Calcd for C92H116N4O16·0.5 CH3CN·0.5 CH2Cl2 C 70.32, H
7.48, N 3.84%; found C 70.40, H 7.39, N 4.00%.
Compounds 2a–d were prepared following the known procedure
with small modification. To a solution of resorcinarene 127
(1 mmol), formaldehyde (37%, 0.75 ml, 10 mmol) and glacial acetic
acid (0.5 ml) in methanol (5 ml), amine (5 mmol) was added.
After 24 h at room temperature the precipitate was filtered off,
washed with methanol and dried. All physicochemical properties
of compounds 2a,b, e and 2c,d, concur with published data.10,12
Procedure for N-substitution (steps b and c)
Method A. To a solution of tetrabezoxazine 2 (0.1 mmol) in
1 ml of methanol, ethyl nitroacteate was added (100 ll). The
solution was stirred for 2h. During that time the suspension
dissolves and TLC shows substrate spot disappearance. To that
solution 50 ll of water was added and 200 mg of Boc2O. The
reaction was stirred overnight. The precipitate was filtered out
and recrystallized from CH2Cl2/MeCN to give product 4. The
filtrate was checked by TLC for the presence of product, and
if the product was detected in noticeable amount the filtrate
was subjected to column chromatography on silica gel (eluent
usually hexane:ethyl acetate 8:2). However, usually the product
after chromatography was still contaminated with traces of
ethyl nitroacetate, so recrystallization/precipitation was required
to obtain analytically pure products. The additional portions
obtained by chromatography did not exceed 10% yield.
Method B. To a solution of tetrabenzoxazine 2 (0.1 mmol)
in 1 ml of methanol, ethyl nitroacetate was added (100 ll). The
solution was stirred for 2 h. During that time the suspension
dissolves and TLC shows substrate spot disappearance. The
solution was then evaporated at reduced pressure (30 ◦C, water
bath) and vacuum dried. The resulting oil was washed with hexane
(3 × 2 ml) and dried again. The oil was dissolved in 2 ml of
dry THF and then 60 ll of tert-butyl isocyanate was added.
The reaction was stirred overnight. The mixture was then treated
with 1 ml of methanol in order to decompose the remaining
isocyanate and evaporated. Crude product 4 was purified by
column chromatography on silica gel using hexane–AcOEt (8 : 2)
and additionally recrystallized/precipitated from CH2Cl2/MeCN.
4a. Method A, yield 56%. dH (500 MHz, CDCl3, TMS, 303 K):
0.96 (d, J = 10.0 Hz, 12 H, Ha), 0.99 (d, J = 10.0 Hz, 12 H, Ha),
1.49 (bs, 4 H, Hb), 2.08 (bs, 20 H, Hc, Hacetyl), 4.10–5.20 (bm, 20H,
Hf, Hd, Hg), 7.20–7.34 (m, 24 H, Har), 8.68 (bs, 4 H, OH), 11.40 (bs,
4f. Method B, yield 63%. dH (500 MHz, CDCl3, TMS, 303 K):
0.94 (d, J = 6.6 Hz, 12 H, Ha), 0.98 (d, J = 6.6 Hz, 12 H, Ha), 1.08
(s, 36 H, C(CH3)3), 1.53 (q, J = 6.6 Hz, 4H, Hb), 2.11 (bs, 8H, Hc),
4.18 (s, 4 H, NH), 4.35–4.95 (bm, 20H, Hf, Hd, Hg), 7.15 (bd, 8H,
J = 7.0 Hz, Har), 7.22–7.29 (m, 16H, Har), 8.77 (bs, 4 H, OH), 12.04
(bs, 4 H, OH). dC (125 MHz, CDCl3, 303 K): 22.79, 22.88, 26.15,
29.08, 31.78, 42.83, 51.09, 51.77, 112.67, 123.35, 124.58, 126.38,
127.40, 128.85, 137.43, 150.59, 151.64, 159.62. dN −273.8 ppm
(NH). ESI MS m/z Calcd for [C96H128N8O12Na]+ 1607.9; found
1608.0, isotope profile agrees.
4g. Method B, yield 25%. dH (500 MHz, CDCl3, TMS, 303 K):
0.97 (t, J = 7.1 Hz, 12 H, CH2-CH3), 1.11 (s, 36 H, C(CH3)3),
1366 | Org. Biomol. Chem., 2007, 5, 1358–1368
This journal is
The Royal Society of Chemistry 2007
©

View Article Online
2.28 (m, 8H, CH2-CH3), 4.22 (s, 4 H, NH), 4.36 (bt,4 H, Hd,),
4.43–4.95 (bm, 16H, Hf, Hg), 7.18–7.32 (m, 24H, Har), 8.82 (bs,
4 H, OH), 12.04 (bs, 4 H, OH). dC (125 MHz, CDCl3, 303 K):
12.71, 26.72, 29.11, 36.33, 42.56 (br), 51.09, 51.69, 112.66, 122.90,
124.42, 124.55, 126.38, 127.41, 128.87, 137.41, 150.65, 151.78,
159.63. ESI MS m/z Calcd for [C88H112N8O12Na]+ 1495.8; found
1495.9, isotope profile agrees. Anal. Calcd for C88H112N8O12·0.7
CH3CN·0.3 CH2Cl2 C 70.50, H 7.56, N 7.97%; found C 70.48, H
7.35, N 8.07%.
4h. Method B, yield 54%. dH (500 MHz, CDCl3, TMS, 303 K):
0.93–0.96 (m, 36 H, CH2-CH3, Ha), 1.29 (m, 4H, Hb), 1.32 (s,
36 H, C(CH3)3), 1.26–1.37 (m, 12 H, CH2, Hb), 1.47–1.53 (m,
16 H, CH2), 2.06 (bs, 8H, Hc), 3.13 (bs, 4H, N-CH2), 3.32 (bs,
4H, N-CH2), 4.24 (bm, 8 H, NH, Hf1), 4.38 (d, J = 14.6 Hz, 4 H,
Hf2), 4.53 (t, J = 7.9 Hz, 4 H, Hd), 7.15 (s, 4H, Har), 8.71 (bs, 4 H,
OH), 11.94 (bs, 4 H, OH). dC (125 MHz, CDCl3, 303 K): 13.87,
20.29, 22.82, 22.87, 26.17, 29.54, 30.08, 31.85, 40.73, 43.06, 47.48,
51.08, 112.66, 123.28, 124.46, 150.70, 151.65, 159.28. ESI MS m/z
Calcd for [C84H136N8O12Na]+ 1472.0; found 1472.0, isotope profile
agrees.
4i. Method B, yield 48%. [a]D = +137.5 (c 1.13, CHCl3). Main
diastereoconformer (86%): dH (500 MHz, CDCl3, TMS, 303 K):
0.91 (m, 48 H, Ha, Hj), 0.97 (d, J = 6.4 Hz, 12 H, Ha), 1.54 (m, 4H,
Hb), 1.62 (bs, 12H, Hh), 2.11 (m, 8H, Hc), 3.94 (s, 4 H, NH), 4.45
(bs, 4H, Hf1), 4.59 (bt, 4 H, Hd), 4.65 (d, J = 14.6 Hz, 4H, Hf2),
5.35 (q, J = 6.8 Hz, 4H, Hg), 7.01 (m, 8H, Hi), 7.21–7.25 (m, 16H,
He, Hi), 8.75 (bs, 4 H, OH), 11.90 (bs, 4 H, OH). dC (125 MHz,
CDCl3, 303 K): 15.21, 22.65, 22.82, 26.35, 28.92, 31.79, 42.64,
43.57, 50.90, 56.02, 112.92, 123.42, 124.64, 126.25, 127.39, 128,67,
141.38, 150.45, 151.51, 158.49. dN −269.1 ppm (NH). ESI MS m/z
Calcd for [C100H136N8O12Na]+ 1664.0, found 1664.0, isotope profile
agrees. Anal. Calcd for C100H136N8O12 C 73.14, H 8.35, N 6.82%;
found C 73.02, H 8.54, N 6.84%. Minor diastereoconformer (14%,
well separated signals only) dH (500 MHz, CDCl3, TMS):1.49 (m,
4H, Hb, Hh), 1.90(bm, 8H, Hc), 4.02 (s, 4 H, NH), 8.77 (bs, 4 H,
OH), 12.00 (bs, 4 H, OH). dC (125 MHz, CDCl3, TMS):22.50,
22.97, 29.03, 56.55, 124.44, 126.25, 128.86, 158.75.
Frames were measured at 0.5◦ intervals with a counting time of
30 se. The data were corrected for Lorentz and polarization effects.
Empirical correction for absorption was applied. Data reduction
and analysis were carried out with the Oxford Diffraction pro-
grams. The structure was solved by direct methods and refined
using SHELXL (X-Seed28 interface). The refinement was based on
F2 for all reflections except those with very negative F2. Weighted
R factors wR and all goodness-of-fit S values are based on F2.
The non-hydrogen atoms were refined with anisotropic thermal
parameters, the H atoms attached to carbon atoms were positioned
geometrically. The OH hydrogen atoms were located from the
Fourier map and then refined with restraints on the bond lengths
only.
Crystal data for 4j: C94.69H126.10N8.65O13.38 (4j·(CH3CN)0.65
·
(CH2Cl2)0.35), M = 1599.53, tetragonal, space group I4 (no. 79),
3
˚
˚
a = b = 23.3843(6), c = 8.2561(2) A, V = 4514.6(2) A , Z = 2, Dc =
1.177 g cm−3, F000 = 1724, MoKa radiation, k = 0.71073 A, T =
˚
173(2)K, 2hmax = 57.4◦, 21498 reflections collected, 4767 unique
(Rint = 0.0378). Final GooF = 0.776, R1 = 0.0468, wR2= 0.1182,
R indices based on 1993 reflections with I > 2r(I) (refinement on
F2), 279 parameters, 3 restraints. Lp and absorption corrections
applied, l = 0.078 mm−1. Absolute structure parameter = 1.2(16).
CCDC reference number 635229. For crystallographic data in
CIF or other electronic format see DOI: 10.1039/b701451a
Acknowledgements
This work was supported by the State Committee for Scientific
Research (Project # N204 086 31/2028). I am grateful to
Prof. Sławomir Szyman´ski for his valuable insights concerning
NMR experiments and data interpretation, Michał Gałęzowski
for SPARTAN calculations and Dr Dorota Gryko for HPLC
experiments. The X-ray measurements were undertaken in the
Crystallographic Unit of the Physical Chemistry Lab. at the
Chemistry Department of the University of Warsaw
4j. Method B, yield 30%. [a]D = +145.3 (c 1.14, CHCl3). Main
diastereoconformer (84%): dH (500 MHz, CDCl3, TMS, 303 K):
0.95 (m, 48 H, Ha, Hj), 1.64 (bs, 12H, Hh), 2.30 (bs, 8H, Hb),
3.96 (s, 4 H, NH), 4.37–4.68 (m, 12H, Hf1.2, Hd), 5.35 (q, J =
6.0 Hz, 4H, Hg), 7.11–7.26 (m, 24 H, Har), 8.79 (bs, 4 H, OH),
11.88 (bs, 4 H, OH). dC (125 MHz, CDCl3, 303 K): 12.47,
15.22, 26.69, 28.95, 29.06, 36.30, 43.45, 50.88, 55.87, 112.83,
122.88, 124.55, 126.51, 127.40, 128.72, 141.43, 150.58, 151.68,
158.48. ESI MS m/z Calcd for [C92H120N8O12H]+ 1529.9; found
1529.9, isotope profile agrees. Anal. Calcd for C92H120N8O12·0.6
CH3CN·0.4 CH2Cl2 C 70.77, H 7.78, N 7.58%; found C 70.92, H
7.80, N 7.63%. Minor diastereoconformer (16%, well separated
signals only) dH (500 MHz, CDCl3, TMS): 1.51 (m, 4H, Hh), 4.04
(s, 4 H, NH), 8.91 (bs, 4 H, OH), 12.00 (bs, 4 H, OH). dC (125 MHz,
CDCl3, TMS): 29.06, 56.50, 124.36, 126.30, 128.86, 158.70.
References
1 Y. Fukushi, Nippon Nogei Kagaku Kaishi, 1998, 72, 1345–1351; Y.
Fukushi, K. Shigematsu, J. Mizutani and S. Tahara, Tetrahedron Lett.,
1996, 37, 4737–4740; X. Borde, C. Nugier-Chauvin, H. Noiret and
N. Patin, Tetrahedron: Asymmetry, 1998, 9, 1087–1090; M. J. Ferreiro,
S. K. Latypov, E. Quinoa and R. Riguera, Tetrahedron: Asymmetry,
1997, 8, 1015–1018; M. J. Ferreiro, S. K. Latypov, E. Quinoa and R.
Riguera, Tetrahedron: Asymmetry, 1996, 7, 2195–2198.
2 A. V. Malkov and P. Kocovsky, Curr. Org. Chem., 2003, 7, 1737–1757;
H. U. Blaser and C. Malan, Adv. Synth. Catal., 2003, 345, 103–151; B.
Pugin, F. Spindler, H. Steiner, M Studer and M. Nakajima, Yakugaku
Zasshi, 2000, 120, 68–75; T. T. L Au-Yeung and A. S. C. Chan, Coord.
Chem. Rev., 2004, 248, 2151–2164.
3 T. Focken, G. Raabe and C. Bolm, Tetrahedron: Asymmetry, 2004, 15,
1693–1706; R. C. J Atkinson, V. C. Gibson and N. J. Long, Chem. Soc.
Rev., 2004, 33, 313–328.
4 A. D. Cort, J. I. M Murua, C. Pasquini, M. Pons and L. Schiaffino,
Chem.–Eur. J., 2004, 10, 3301–3307.
5 F. Fabris, L. Pellizzaro, C. Zonta and O. De Lucchi, Eur. J. Org. Chem.,
2007, 283–291.
6 A. D. Cort, L. Mandolini, C. Pasquini and L. Schiaffino, New J. Chem.,
2004, 28, 1198–1199.
X-Ray crystallographic structure determination of 4j
7 Cycloenantiomerism was first discussed in the case of cyclopeptides and
catenanes: V. Prelog and H. Gerlach, Helv. Chim. Acta, 1964, 47, 2288–
2294; M. D. G. Schill, Catenanes, Rotaxanes, and Knots, Academic
Press, New York, 1971, pp. 11–15; C. Reuter, R. Schmieder and F.
Vogtle, Pure Appl. Chem., 2000, 72, 2233–2241.
The diffraction quality crystals were grown from CH2Cl2/MeCN
solution. The measurement was performed on a KM4CCD j-axis
diffractometer with graphite-monochromated MoKa radiation.
The crystal was positioned at 62 mm from the CCD camera. 1500
This journal is
The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 1358–1368 | 1367
©

View Article Online
8 M. D. Singh, J. Siegel, S. E. Biali and K. Mislow, J. Am. Chem. Soc.,
1987, 109, 3397–3402; K. Mislow, Chimia, 1986, 40, 395–402.
9 V. Bo¨hmer, D. Kraft and M. Tabatabai, J. Inclusion Phenom. Mol.
Recognit. Chem., 1994, 19, 17–39.
10 R. Arnecke, V. Bo¨hmer, E. F. Paulus and W. Vogt, J. Am. Chem. Soc.,
1995, 117, 3286–3287.
11 P. C. B Page, H. Heaney and E. P. Sampler, J. Am. Chem. Soc., 1999,
121, 6751–6752; M. Klaes, C. Agena, M. Kohler, M. Inoue, T. Wada,
Y. Inoue and J. Mattay, Eur. J. Org. Chem., 2003, 1404–1409.
12 W. Iwanek and J. Mattay, Liebigs Ann. Chem., 1995, 1463–1466; M. T. El
Gihani, H. Heaney and A. M. Z. Slawin, Tetrahedron Lett., 1995, 36,
4905–4908; R. Arnecke, V. Bo¨hmer, S. Friebe, S. Gebauer, G. J. Krauss,
I. Thondorf and W. Vogt, Tetrahedron Lett., 1995, 36, 6221–6224; A.
Szumna, M. Gorski and O. Lukin, Tetrahedron Lett., 2005, 46, 7423–
7426.
21 B. R. Buckley, J. Y. Boxhall, P. C. B. Page, Y. Chan, M. R. J. Elsegood,
H. Heaney, K. E. Holmes, M. J. McIldowie, V. McKee, M. J. McGrath,
M. Mocerino, A. M. Poulton, E. P. Sampler, B. W. Skelton and A. H.
White, Eur. J. Org. Chem., 2006, 5117–5134.
22 C. L. Perrin and T. J. Dwyer, Chem. Rev., 1990, 90, 935–967.
23 J. Sartorius and H. J Schneider, Chem.–Eur. J., 1996, 2, 1446–1452.
24 C. Schiel, G. A. Hembury, V. V. Borovkov, M. Klaes, C. Agena, T.
Wada, S. Grimme, Y. Inoue and J. Mattay, J. Org. Chem., 2006, 71,
976–982.
25 G. S. Kottas, L. I. Clarke, D. Horinek and J. Michl, Chem. Rev., 2005,
105, 1281–1376.
26 M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb,
J. R. Cheeseman, J. A. Montgomery, Jr., T. Vreven, K. N. Kudin, J. C.
Burant, J. M. Millam, S. S. Iyengar, J. Tomasi, V. Barone, B. Mennucci,
M. Cossi, G. Scalmani, N. Rega, G. A. Petersson, H. Nakatsuji, M.
Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T.
Nakajima, Y. Honda, O. Kitao, H. Nakai, M. Klene, X. Li, J. E.
Knox, H. P. Hratchian, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo,
R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi,
C. Pomelli, J. Ochterski, P. Y. Ayala, K. Morokuma, G. A. Voth, P.
Salvador, J. J. Dannenberg, V. G. Zakrzewski, S. Dapprich, A. D.
Daniels, M. C. Strain, O. Farkas, D. K. Malick, A. D. Rabuck, K.
Raghavachari, J. B. Foresman, J. V. Ortiz, Q. Cui, A. G. Baboul, S.
Clifford, J. Cioslowski, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz,
I. Komaromi, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y.
Peng, A. Nanayakkara, M. Challacombe, P. M. W. Gill, B. G. Johnson,
W. Chen, M. W. Wong, C. Gonzalez and J. A. Pople, GAUSSIAN 03
(Revision B.05), Gaussian, Inc., Pittsburgh, PA, 2003.
13 O. Trapp, S. Caccamese, C. Schmidt, V. Bo¨hmer and V. Schurig,
Tetrahedron: Asymmetry, 2001, 12, 1395–1398.
14 D. M. Rudkevich, G. Hilmersson and J. Rebek, J. Am. Chem. Soc.,
1997, 119, 9911–9912.
15 S. Saito, C. Nuckolls and J. Rebek, J. Am. Chem. Soc., 2000, 122,
9628–9630.
16 A. Shivanyuk, K. Rissanen, S. K. Korner, D. M. Rudkevich and J.
Rebek, Helv. Chim. Acta, 2000, 831778–1790.
17 O. Hayashida, S. Matsumoto and I. Hamachi, Chem. Lett., 2005, 34,
1276–1277; O. Hayashida, J. Ito, S. Matsumoto and I. Hamachi, Org.
Biomol. Chem., 2005, 3, 654–660.
18 L. Frkanec, A. Visnjevac, B. Kojic-Prodic and M Zinic, Chem.–Eur. J.,
2000, 6, 442–453.
19 C. Schmidt, E. F. Paulus, V. Bo¨hmer and W. Vogt, New J. Chem., 2000,
24, 123–125.
20 S. V. Chankeshwara and A. K. Chakraborti, Org. Lett., 2006, 8, 3259–
3262.
27 L. M. Tunstad, J. A. Tucker, E. Dalcanale, J. Weiser, J. A. Bryant, J. C.
Sherman, R. C. Helgeson, C. B. Knobler and D. J. Cram, J. Org. Chem.,
1989, 54, 1305–1312.
28 L. J. Barbour, J. Supramol. Chem., 2001, 1, 189–191.
1368 | Org. Biomol. Chem., 2007, 5, 1358–1368
This journal is
The Royal Society of Chemistry 2007
©