Optically active dioxatetraazamacrocycles: Chemoenzymatic syntheses and applications in chiral anion recognition

Article abstract of DOI:10.1002/1521-3765(20000915)6:18<3331::AID-CHEM3331>3.0.CO;2-8

Two new C₂ and D₂ symmetrical dioxatetraaza 18-membered macrocycles [(R,R)-1 and (S,S,S,S)-2] are efficiently synthesized in enantiomerically pure forms by a chemoenzymatic method starting from (±)-trans-cyclohexane-1,2-diamine. The

Full text of DOI:10.1002/1521-3765(20000915)6:18<3331::AID-CHEM3331>3.0.CO;2-8

FULL PAPER
Optically Active Dioxatetraazamacrocycles:
Chemoenzymatic Syntheses and Applications in Chiral Anion Recognition
Ignacio Alfonso, Francisca Rebolledo, and Vicente Gotor*[a]
Â
Dedicated to Professor Jose Elguero on the occasion of his 65th birthday
Abstract: Two new C2 and D2 symmet-
rical dioxatetraaza 18-membered mac-
rocycles [(R,R)-1 and (S,S,S,S)-2] are
efficiently synthesized in enantiomeri-
cally pure forms by a chemoenzymatic
method starting from (Æ)-trans-cyclo-
hexane-1,2-diamine. The protonation
constants and the binding constants with
different chiral dicarboxylates are deter-
mined in aqueous solution by means of
pH-metric titrations. The triprotonated
form of (S,S,S,S)-2 shows moderate
enantioselectivity with malate and tar-
tonated (R,R)-1 and N-acetyl aspartate,
the complex with the d-enantiomer
being 0.92kcalmol À1 more stable than
its diastereomeric counterpart. Despite
the lack of enantioselectivity of tri- and
tetraprotonated (R,R)-1 for the tartrate
anion, a very good diastereopreference
for meso-tartrate is found. All these
experimental results allow us to propose
a model for the host ± guest structure
based on coulombic interactions and
hydrogen bonds.
trate anions (DDG ˆ 0.62and 0.66 kcal
molÀ1, respectively), being the strongest
binding observed in both cases with the
l enantiomer. Good enantiomeric dis-
crimination is obtained with tetrapro-
Keywords: chiral anion recognition
´ enantiomeric resolution ´ host ±
guest chemistry
´ macrocycles ´
molecular recognition
phosphorylase[6], and enolase[7] activities in comparison with
the normal enzymatic catalysis.
Introduction
Oxaazamacrocycles are attractive molecules with a variety of
applications in the fields of supramolecular chemistry and
catalysis.[1] The presence of two kinds of heteroatoms (nitro-
gen and oxygen) in the macrocyclic ring confer hybrid
properties, between those of azacrowns and crown ethers,
on these structures.[2] Thus, depending on structural factors
like size of the ring or number and disposition of the
heteroatoms, these systems are able to form kinetically and
thermodynamically stable complexes with different metals.[3]
Furthermore, when protonated, they are excellent anion
binders.[4] For instance, some of these compounds are
receptors for biologically important anions like nucleotides
and nucleic acids. The complexation affects the reactivity of
the anion, leading to interesting applications in supramolec-
ular catalysis. Therefore, 24- to 36-membered-ring oxaaza-
macrocycles have been extensively studied for their ATPase,[5]
However, the design and synthesis of selective receptors for
optically active polyanions still constitutes a difficult task
facing chemists. Although cyclodextrins have proved to be
excellent ligands for the recognition of helicity,[8] only low
selectivities have been obtained with compounds bearing
chiral centers.[9] Recently, sapphyrin-based receptors[10] have
been successfully used for the recognition and transport of
aspartate and glutamate derivatives; these are the first
receptors capable of the selective recognition of dianionic
dicarboxylates with chiral centers.
Taking into account the interest and scarcity of suitable
optically active macrocycles that act as selective receptors of
chiral anions,[11] in this work we aimed at the synthesis of
optically active oxaazamacrocycles 1 and 2, with C2 and D2
symmetry, respectively, by a chemoenzymatic method. In
addition, we studied their protonation and their chiral anion
recognition properties in aqueous solution, since this is the
natural environment of biologically important anions.
[a] Prof. Dr. V. Gotor, Dr. I. Alfonso, Dr. F. Rebolledo
Â
Â
Departamento de Química Organica e Inorganica
Universidad de Oviedo
Results and Discussion
33071 Oviedo (Spain)
Fax : (‡34)98-510-3448
Design and synthesis: Structural complementarity in molec-
ular recognition is crucial for the success of a given
receptor.[12] Coulombic attractions and hydrogen-bond for-
E-mail: vgs@sauron.quimica.uniovi.es
Supporting information for this article is available on the WWW under
http://www.wiley-vch.de/home/chemistry or from the author.
Chem.Eur.J. 2000, 6, No. 18
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3331 $ 17.50+.50/0
3331

V. Gotor et al.
FULL PAPER
dioxane as solvent and with the molar ratio diamine:diester
varied depending on the required conversion. Thus, when
equimolecular amounts of (Æ)-3 and 4 were used, a mixture
containing the double aminolysis compound bis(amidoester)
(R,R)-5 and the remaining (S,S)-3 was obtained (Scheme 1),
mation play a dominant role in the anion complexation of
polyammonium macrocycles.[13] The strongest binding usually
occurs with the fully protonated form of the azamacrocycle
and when substrate and receptor are preorganized into
complementary shapes. In addition, compounds containing
an optically active trans-cyclohexane-1,2-diamine moiety have
proved to be very useful in both asymmetric synthesis[14] and
enantiomeric and diastereomeric recognition of peptides.[15]
These were the reasons we believed compounds 1 and 2 would
be good candidates for the recognition of optically active a,w-
dicarboxylates. Both oxaazamacrocycles are 18-membered
rings, a convenient size for the complexation of 1,2-dicarbox-
ylates.[16] They have two ether units separating the N C C N
À À À
fragments, which increase the pH values for full protonation
and, furthermore, may cooperate in the binding as hydrogen-
bond acceptors. Compound 1 has C2 symmetry, and 2
possesses three orthogonal binary axes to give it overall D2
symmetry.
Scheme 1. Enzymatic resolution of (Æ)-3. a) CAL, 1,4-dioxane, 308C;
b) HCl (g).
the product of the monoacylation of the diamine being
undetected. After the elimination of the lipase from the
medium, the remaining diamine (S,S)-3 was transformed into
the corresponding dihydrochloride by treatment with dry
hydrogen chloride in order to facilitate separation and
isolation of the compounds. The enantiomeric excess of
(R,R)-5 was determined by chiral HPLC (see the Exper-
imental Section), whereas the diamine (S,S)-3 was determined
as previously described.[18] As is indicated in Table 1 (entry 1),
Racemic trans-cyclohexane-1,2-diamine [(Æ)-3] is an inter-
esting substrate for lipase-catalyzed kinetic resolutions. The
presence of two amine groups of the same configuration
allows us to carry out a sequential kinetic resolution.[17]
Recently, we applied this methodology to the resolution of
(Æ)-3 with dimethyl malonate using lipase B from Candida
antarctica (CAL) as catalyst.[18] Now, with the structure of
azamacrocycles 1 and 2 in mind, we focused on extending this
biocatalytic process to the resolution of (Æ)-3 with dimethyl
3-oxapentanedioate (4). Reactions were carried out in 1,4-
Table 1. Enzymatic resolution of (Æ)-3.
Molar ratio (Æ)-3:4
t [h]
ee (S,S)-3 [%]
ee (R,R)-6 [%]
1:1
1:1.3
1:1.5
9
15
39
92
97
> 99
> 99
> 99
> 99
Abstract in Spanish: Se han sintetizado dos nuevos dioxate-
traazamacrociclos de 18 eslabones [(R,R)-1 y (S,S,S,S)-2]
Â
enantiomericamente puros y con simetría C2 y D2, a partir de la
Â
(Æ)-trans-ciclohexano-1,2-diamina mediante un metodo qui-
Â
Â
mioenzimatico.Adema s, se han determinado tanto las cons-
enantiomeric excesses were very high for both compounds,
(R,R)-5 being in fact enantiopure. In order to improve the ee
of (S,S)-3, the conversion degree of the enzymatic reaction has
to be improved and, consequently, an excesss of diester 4 must
be used. Enantiopure (S,S)-3 was obtained when a 1:1.5 molar
ratio of 3:4 was used for a longer reaction time. Large
amounts of monoacylation product (ca. 30%) and enantio-
pure (R,R)-5 were also isolated in this process. These
compounds [(R,R)-5 and (S,S)-3] have been used in this work
for the syntheses of the oxaazamacrocycles (R,R)-1 and
(S,S,S,S)-2, respectively.
Â
Â
tantes de protonacion como las de complejacion con diferentes
dicarboxilatos quirales en disolucion acuosa mediante valora-
ciones potenciometricas.La forma triprotonada de (S,S,S,S)- 2
Â
Â
muestra una enantioselectividad moderada con los aniones
malato y tartrato (DDG ˆ 0.62 y 0.66 kcalmolÀ1, respectiva-
Â
Â
mente), apareciendo en ambos casos la interaccion mas fuerte
Â
Â
con el enantiomero l.Se obtiene una buena discriminacio n
Â
enantiomerica con (R,R)-1 en su forma tetraprotonada y N-
Â
acetil aspartato, siendo el complejo con el enantiomero d
0.92 kcalmolÀ1 mas estable que con su correspondiente enan-
Â
Â
tiomero.A pesar de la ausencia de enantioselectividad de
The oxaazamacrocycle (R,R)-1 was synthesized from
(R,R)-5 following a modification of the Richmann ± At-
kins[19] procedure, which mainly involves the coupling of a
tosylated polyamine (R,R)-8 with a doubly electrophilic
reagent in the presence of a base (Scheme 2). The ester
functions of (R,R)-5 were quantitatively transformed into
amides by conventional ammonolysis (ammonia in metha-
Â
(R,R)-1 tri- y tetraprotonados con el anion tartrato, se
encuentra una diastereopreferencia importante hacia el meso-
tartrato.Todos estos resultados experimentales nos permiten
proponer un modelo para la estructura supramolecular, basado
Â
en interacciones culombianas y enlaces por puente de hidro-
geno.
3332
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3332 $ 17.50+.50/0
Chem.Eur.J. 2000, 6, No. 18

Dioxatetraazamacrocycles
3331±3338
manufactured by the addition of concentrated HCl to a
methanolic solution of (R,R)-1 and subsequent evaporation to
dryness. The resulting solid was easily recrystallized in ethanol
to give (R,R)-1 ´ 4HCl with a 56% overall yield from (R,R)-5.
For the synthesis of the macrocycle (S,S,S,S)-2, we started
from the enantiopure (S,S)-3 ´ 2HCl (see Table 1, entry 3),
following a modification of the Biernat and Luboch[20]
procedure. It is well known that the formation of nine-
membered rings is kinetically disfavored[21] and that 18-
membered rings are the ones most easily obtained by means
of the Richman ± Atkins procedure.[19b] With these antece-
dents we envisioned the possibility of preparing the macro-
cycle (S,S,S,S)-2 by a [2‡2] cyclization reaction in which two
molecules of the nucleophile react with two molecules of the
electrophile (Scheme 3). Thus, the reaction of the bis(sulfon-
Scheme 2. Synthesis of (R,R)-1. a) NH3 in MeOH; b) BH3 ´ THF; c) HCl
(6n); d) TsCl, K2CO3, THF, H2O; e) Cs2CO3, CH3CN, MsOCH2CH2OMs;
f) HBr, PhOH; g) NaOH (aq)/CH2Cl2; h) HCl in MeOH.
nol). Subsequent reduction (BH3/THF) of the tetraamide
(R,R)-6 led to the tetraamine (R,R)-7, which was isolated and
used in the next step as its tetrahydrochloride. Tosylation of
(R,R)-7 ´ 4HCl with tosyl chloride was efficiently carried out
in a biphasic system formed by tetrahydrofuran and aqueous
potassium carbonate.
Scheme 3. Synthesis of (S,S,S,S)-2. a) TsCl, K2CO3, THF, H2O; b) Cs2CO3,
CH3CN, MsOCH2CH2OCH2CH2OMs; c) HBr, PhOH; d) NaOH (aq)/
CH2Cl2; e) HCl in MeOH.
The coupling of (R,R)-8 with ethylene glycol dimesylate
was accomplished with an excess of cesium carbonate in
refluxing acetonitrile (Scheme 2). Under these conditions, the
tetratosylated oxaazamacrocycle (R,R)-9 was obtained with
87% yield after purification by flash chromatography. The
cyclic structure was unambiguosly demonstrated by fast atom
bombardment (FAB) mass spectrometry: the spectrum
showed peaks at m/z values of 931 and 953, corresponding
amide) (S,S)-10, obtained from (S,S)-3 ´ 2HCl,[22] with dieth-
ylene glycol dimesylate led to the tosylated oxaazamacrocycle
(S,S,S,S)-11 in 80% yield after chromatographic purification.
The reaction was carried out in refluxing acetonitrile and with
an excess of alkaline carbonate as a base. No significant
differences were observed when cesium or potassium salts
were used, but with the former, the reaction time was shorter.
Compound (S,S,S,S)-11 showed a clear melting point (200 ±
2028C) and a single spot on TLC in a variety of eluents. The
FAB mass spectrum showed peaks at m/z values of 985 and
‡
‡
to [M‡H] and [M‡Na] , respectively. Hydrolysis of the
sulfonamide functions of (R,R)-9 with 48% HBr and an
excess of phenol yielded the desired dioxatetraazamacrocycle
(R,R)-1, isolated as its tetrahydrobromide. The unprotonated
species can be easily obtained as a colorless oil by addition of
4n sodium hydroxyde and extraction with dichloromethane.
Considering that the asymmetric centers do not participate in
any reaction step, no racemization is expected in this syn-
‡
‡
1007, corresponding to [M‡H] and [M‡Na] , respectively.
These facts demonstrated again the 18-membered cyclic
structure for the product and revealed the success of the
[2‡2] cyclization strategy in this synthesis. Hydrolysis of the
resulting sulfonamide (S,S,S,S)-11 was carried out under the
conditions already described for (R,R)-10, yielding compound
(S,S,S,S)-2 ´ 4HBr, which was also converted into the corre-
sponding tetrahydrochloride salt to facilitate its purification.
Again the analysis of (S,S,S,S)-2 ´ 4HCl by NMR spectroscopy
revealed the expected symmetry. For example, the 13C NMR
spectrum showed only five signals corresponding to an
effective D2 symmetry in solution, demonstrating that epi-
merization procceses had not taken place.
1
thesis. In confirmation, H and 13C NMR spectra of (R,R)-1
and of its precursors [(R,R)-7 and (R,R)-8] showed the
expected signals for an effective C2 symmetry (see exper-
imental data). These facts demonstrate that in the above
synthetic routes epimerization of the chiral centers has not
taken place, since epimerization would lead to the formation
of cis/trans diastereomeric mixtures and thus to an increase of
the number of signals in the spectra.
In order to facilitate the purification of the macrocyclic salt
and for the potentiometric measurements, (R,R)-1 ´ 4HCl was
Chem.Eur.J. 2000, 6, No. 18
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3333 $ 17.50+.50/0
3333

V. Gotor et al.
FULL PAPER
Table 3. Logarithms of stability constants[a] for the interaction of (R,R)-1
and (S,S,S,S)-2 with chiral dicarboxylate anions, determined at 298 K in
aqueous NMe4Cl (0.1 moldmÀ3).
Protonation behavior: For the study of anion recognition by
oxaazamacrocycles in aqueous solution, a preliminary knowl-
edge of the protonation constants is necessary. The behavior
of (R,R)-1 and (S,S,S,S)-2 towards protonation has been
studied in 0.1 moldmÀ3 NMe4Cl solution at 298 K in the pH
range 2± 11.5. The values of the logarithm of the basicity
constant for each protonation step of these macrocycles are
shown in Table 2, together with the one previously reported
n[b] ˆ 4
n[b] ˆ 3
Macrocycle Anion
logKs
DlogKs[c] logKs
DlogKs[c]
(R,R)-1
(R,R)-1
(R,R)-1
(R,R)-1
(R,R)-1
(R,R)-1
(R,R)-1
(R,R)-1
d-tartrate
l-tartrate
meso-tartrate 3.92(1)
d-malate
l-malate
N-Ac-d-Asp
N-Ac-l-Asp
N-Ac-d-Glu
N-Ac-l-Glu
(R)-Me-succ
(S)-Me-succ
d-tartrate
l-tartrate
meso-tartrate 2.69(2)
d-malate
l-malate
N-Ac-d-Asp
N-Ac-l-Asp
N-Ac-d-Glu
N-Ac-l-Glu
(R)-Me-succ
(S)-Me-succ
2.59(1)[d]
2.56(1)
1.92(1)
1.85(1)
2.73(1)
2.14(1)
2.21(1)
3.40(1)
3.26(1)
3.12(2)
3.12(2)
4.31(2)
4.20(2)
1.68(3)
2.16(1)
1.94(2)
1.91(3)
2.36(2)
2.49(3)
2.53(3)
2.72(3)
2.80(3)
3.00(2)
3.01(2)
0.03
0.07
2.96(2)
±
4.21(1)
3.54(3)
3.52(3)
±
5.43(2)
5.42(2)
±
n.a.[e]
0.67
À 0.07
0.14
Table 2. Logarithms of the protonation constants measured by pH-metry
in NMe4Cl (0.1 moldmÀ3) at 298 K.
(R,R)-1
(R,R)-1
(R,R)-1
n.a.[e]
0.01[f]
0.00
[b]
trans-18-ane-N4O2
(R,R)-1
(S,S,S,S)-2
logK
Reaction[a]
logK
D
logK
D
D
0.11[f]
À 0.48
L ‡ H ˆ HL
9.36
9.06(3)[c]
7.97(3)
5.53(3)
3.88(4)
10.19(2)
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
(S,S,S,S)-2
HL ‡ H ˆ H2L 8.40
H2L ‡ H ˆ H3L 6.27
H3L ‡ H ˆ H4L 5.23
0.96
2.13
1.04
1.09
2.44
1.65
9.10(1) 1.09
5.29(2) 3.81
4.26(2) 1.03
±
2.47(3)
2.80(3)
À 0.33
À 0.45
À 0.04
À 0.08
À 0.01[f]
[a] Charges are omitted for clarity. [b] 7,16-Dioxa-1,4,11,13-tetraazacy-
clooctadecane; logK taken from ref. [3], measured in 0.1n NaNO3 as
supporting electrolyte. [c] Values in parentheses are standard deviations
on the last significant figure.
±
±
±
±
±
±
for the structurally related trans-18-ane-N4O2 (7,16-dioxa-
1,4,11,13-tetraazacyclooctadecane).[3] In this table we also
present the difference between successive protonation con-
stants for each compound (D), which can be interpreted as the
ease with which a protonated species accepts the next proton:
the higher the D value, the less favorable the protonation
process. The common characteristic of both systems is that the
third protonation step is the most difficult, as expected from
the reported general protonation properties of macrocyclic
polyamines.[23] However, some differences in the value of D
are due to the presence of the cyclohexane. So, when the
second proton is attached to the monoprotonated cyclo-
hexane-1,2-diamine moiety, the D value is increased (see
Table 2). This effect is shown in the fourth step for (R,R)-1
and in the third for (S,S,S,S)-2 compared with trans-18-ane-
N4O2, and is in accordance with the previously reported
protonation constants of ethylenediamine (D ˆ 2.87)[24] and
trans-cyclohexane-1,2-diamine (D ˆ 3.17).[25]
n‡
[a] As defined by the equation: A2À ‡ LHn ˆ [ALHn](nÀ2)‡. [b] Number
of protons accepted by the complex. [c] DlogKs ˆ logKs(D) À logKs(L).
[d] Values in parentheses are standard deviations on the last significant
figure. [e] Not applicable. [f] DlogKs ˆ logKs(R) À logKs(S).
some protonated form of both oxaazamacrocycles, recognition
is less effective with compound (S,S,S,S)-2, probably due to
the more rigid structure and less suitable preorganization of
this receptor. Thus many of the anions form no complexes
‡
with 2 ´ 4H . In contrast, a very good chain selectivity is
observed for (R,R)-1 if binding constants with the acetyl
derivatives of d-aspartate and d-glutamate are compared. Tri-
and, particularly, tetraprotonated (R,R)-1 bind the shorter N-
Ac-d-aspartate more strongly, as would be expected for the
cavity size of the macrocycle.
Concerning enantioselectivity, a general tendency is ob-
served for (S,S,S,S)-2 to be l-selective and (R,R)-1 to be
mainly d-selective, as a consequence of the different absolute
configuration of the oxaazamacrocycles. Two significantly
different binding constants were found for the two diastereo-
Chiral anion recognition: As previously stated, chiral anion
recognition is a nearly unexplored field in supramolecular
chemistry.[1, 4a, 13] There are many polyanions with important
biological activities; for instance, malate or succinate are
intermediates in the citric acid and glyoxylate cycles,[26] and
aspartate and glutamate have been thoroughly studied as
excitatory aminoacid neurotransmitters.[27] We have studied
the complexation of some of these derivatives with the
protonated forms of macrocycles (R,R)-1 and (S,S,S,S)-2.
The stability constants of the complexes of (R,R)-1 and
(S,S,S,S)-2 with some chiral anions were measured in a
0.1 moldmÀ3 NMe4Cl solution at 298 K by pH-metric titra-
tions; they are listed in Table 3. The enantioselectivity is
shown as the difference between the logarithms of the binding
constants obtained for both enantiomers of the anion. Some
general conclusions can be extracted from these experimental
results. Although all the anions form stable complexes with
‡
meric complexes obtained from 2 ´ 3H and tartrate and
malate anions (DlogKs ˆ 0.48 and 0.45, corresponding to
D(DG) ˆ 0.66 and 0.62kcalmol À1), representing a binding
enantioselectivity of ꢀ75:25. The selectivity of (R,R)-1
towards malate and N-Ac-glutamate is manifested in the
‡
absence of stable complexes of 1 ´ 4H with the l enantiomers.
However, this significant result lacks practical applicability
‡
‡
because 1 ´ 4H and 1 ´ 3H species are present together in
solution in a narrow pH range and the latter does not
discriminate enantiomerically. The best result was obtained
‡
with (R,R)-1 ´ 4H and N-Ac-aspartate, for which enantio-
selectivity was quite high (DlogKs ˆ 0.67, DDG ˆ
0.92kcalmol À1), leading to a binding enantioselectivity of
ꢀ82:18, although the presence in solution of the complex with
‡
the less selective 1 ´ 3H would decrease this ratio.
3334
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3334 $ 17.50+.50/0
Chem.Eur.J. 2000, 6, No. 18

Dioxatetraazamacrocycles
3331±3338
In spite of the lack of enantioselectivity shown by (R,R)-1
towards the two enantiomers of the tartrate anion, the
excellent diastereoselectivity (see Table 3) exhibited by this
receptor both in its tri- and tetraprotonated forms is surpris-
ing. Comparison of the logKs values obtained for the
revealed a flat conformation for the cycle, with all the
ethylene groups in a gauche disposition and the ammonium
groups towards the inner macrocyclic cavity. This conforma-
tion settles one lone pair of each oxygen in an endo position to
the macrocyclic ring, leaving the other one free for possible
hydrogen-bond interactions, as in the conformation shown in
Figure 1a. The other two binding points must be due to
‡
complexes of 1 ´ 4H with, for example, d-tartrate and meso-
tartrate gives DDG ˆ 1.8 kcalmolÀ1, meaning a very high
diastereoselective discrimination. This fact, which must be an
effect of the geometry of the complex, was an essential clue to
the host ± guest structure that would allow us to explain the
results.
In order to test the effect of the macrocyclic ring, we
measured the binding constants between all these anions and
(S,S)-cyclohexane-1,2-diamine as a control compound. The
stability constants of the complexes formed are shown in
Table 4. No complexation with tartrate isomers was detected
Table 4. Logarithm of stability constants[a] for the interaction of (S,S)-3 ´
‡
2H with chiral dicarboxylate anions, determined at 298 K in aqueous
NMe4Cl (0.1 moldmÀ3).
Anion
logKs
DlogKs[b]
d-tartrate
±
l-tartrate
±
±
meso-tartrate
d-malate
l-malate
N-Ac-d-Asp
N-Ac-l-Asp
N-Ac-d-Glu
N-Ac-l-Glu
(R)-Me-succ
(S)-Me-succ
±
1.45(6)[c]
1.59(3)
2.07(2)
1.97(2)
2.30(2)
2.32(2)
2.60(1)
2.42(1)
À 0.14
0.10
À 0.02
0.18[d]
Figure 1. Proposed model for the host ± guest structure: a) schematic
representation of the interactions; models for the complexes of (R,R)-1 ´
‡
4H with b) meso- and l-tartrate; c) d- and l-N-Ac-aspartate.
[a] As defined by the equation: A2À ‡ LH2 ˆ [ALH2]. [b] DlogKs ˆ
logKs(D) À logKs(L). [c] Values in parentheses are standard deviations
on the last significant figure. [d] DlogKs ˆ logKs(R) À logKs(S).
2‡
carboxylate ± diammonium coulombic interactions. Concern-
ing the conformation of the anion, previous studies of the
complexation of achiral dicarboxylates with 18-membered-
ring hexaazamacrocycles revealed that they recognize mal-
eate but not fumarate,[16] and a gauche conformation (syn-
type) is favored compared with the anti conformation of the
aliphatic dicarboxylate anions. With all these data in hand, we
propose the model shown in Figure 1a, which allows two
carboxylate ± diammonium interactions and an additional
hydrogen bond with the oxygen of the macrocycle. With this
model, we can explain the diastereomeric preference for the
tartrate anion. The meso derivative can form two cooperative
hydrogen bonds (see Figure 1b), whereas for both d and l
enantiomers only one bond is possible. Besides, the impor-
tance of the hydrogen bond in enantiomeric recognition is
supported by the results obtained with methylsuccinate,
where no enantiodiscrimination was observed (see Table 3).
Furthermore, we used this model to explain the enantiose-
under the experimental conditions (0.1 moldmÀ3 NMe4Cl and
298 K), while the other anions interact with the diprotonated
diamine. It must be pointed out that the higher the hydro-
phobicity of the anion, the larger the binding constant is. In
addition to this, the lack of enantiomeric and diastereomeric
discrimination (DlogKs very close to experimental error)
clearly implies the importance of the crown ring in supra-
molecular chiral recognition.
Another independent proof of the macrocycle ± anion
interaction was obtained from NMR experiments carried
out with the anion that forms the most stable complex (see
Supporting Information). The ROESY spectrum of a 1:4
‡
mixture of (R,R)-1 ´ 4H :(R)-Me-succ shows a cross-peak
between CH2N protons of the macrocycle and CH3 protons of
the anion, evidence supporting the formation of the complex
in solution. Unfortunately, it was not possible to obtain similar
data for the other anions.
‡
lectivity exhibited by 1 ´ 4H towards the aspartate derivative.
Taking into account all these experimental results, we have
proposed a model for the host ± guest complex structure. It is
well known that for high enantioselectivity (or diastereose-
lectivity), a receptor must have three points of interaction
with the substrate, and at least one of them must be
Figure 1c shows the possible structures of the diastereomeric
complexes formed with the pair of enantiomers of this anion.
Both of them can, a priori, achieve the three proposed
interactions, although the relative disposition of the anion in
each is obviously different. In order to give an explanation of
the results obtained with this anion, molecular mechanics
calculations[30] have been undertaken. Figure 2shows the
stereochemically dependent.[28] Molecular dynamic simula-
[29]
tions of tetraprotonated trans-18-ane-N4O2
in solution
Chem.Eur.J. 2000, 6, No. 18
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3335 $ 17.50+.50/0
3335

V. Gotor et al.
FULL PAPER
analyses were performed on a Perkin ± Elmer 240B elemental analyser. 1H
and 13C NMR spectra were obtained with a Bruker AC-300 (1H, 300 MHz,
and 13C, 75.5 MHz), a Bruker AC-200 (1H, 200 MHz, and 13C, 50.3 MHz), or
a Bruker AMX-400 (1H, 400 MHz, and 13C, 100.7 MHz) spectrometer. For
the pH-metric titrations a Metrohm-702titrimeter was used, the reference
electrode was an Ag/AgCl electrode in saturated aqueous KCl, the cell was
thermostated at 298 Æ 0.1 K, the solution stirred, and all measurements
were perfomed under nitrogen. The protonation constants were deter-
mined by titration with 0.1n NaOH of a solution typically containing 10À3
m
of the HCl salt of the oxaazamacrocycle in the presence of Me4NCl (0.1m).
The logKs values of the complexes were determined by titration with 0.1n
NaOH of a solution containing 10À3 m of the HCl salt of the polyamine and
5 Â 10À3 m of the desired dianions in the presence of 0.1m NMe4Cl. All the
measurements with each system were carried out at least twice and the data
analysis was performed with the computer program SUPERQUAD.[34] The
titration curves for each system were treated either as separated entities or
as a single set without significant variations in the stability constants.
Figure 2. MMX force-field minimized structures for the diastereomeric
complexes of (R,R)-1 ´ 4H and d- and l-N-Ac-aspartate.
‡
Dimethyl 3-oxapentanedioate (4): Diglycolyl chloride (5 mL, 40 mmol)
was added dropwise at 08C to a vigorously stirred solution of triethylamine
(12mL) in dry methanol (20 mL). The reaction mixture was allowed to
reach room temperature, then 3n HCl was added and extracted with
dichloromethane (3 Â 30 mL). The organic layers were dried and evapo-
rated to dryness, giving an oily substance which solidified at 08C. The solid
thus obtained was washed with hexane, yielding compound 4 (6.423 g, 94%
yield). M.p. 39 ± 408C; IR (nujol) 1754 cmÀ1; 1H NMR (CDCl3, 200 MHz):
d ˆ 3.77 (s, 6H; CH3), 4.25 (s, 4H; CH2); 13C NMR (CDCl3, 75.5 MHz) d ˆ
optimized structures of both diastereomeric complexes of 1 ´
4H and N-Ac-aspartate enantiomers. MMX force-field
‡
calculations predict a DDHf of 0.66 kcalmolÀ1 favorable to
the d enantiomer, which is in good agreement with the
experimental results (see Supporting Information). Inspec-
tion of the different contributions to the calculated energy
shows a large difference in the torsional energy of both
diastereomeric complexes. This fact and the comparison of
the optimized geometries (see Figure 2) suggest that the
complex with the l enantiomer is less stable owing to greater
folding of the macrocycle upon complexation. Besides, the
distortion of the structure is more important in the ethylene
fragment than in the cyclohexane moiety. This could also
explain the lower binding constants of 2 compared with 1; the
presence of the second six-membered ring in 2 would increase
its rigidity and, therefore, would disfavor the binding with the
dianion. Although entropically driven electrostatic forces
have been identified as the most important contributions in
anion binding,[31] other interactions like hydrogen bonds[32]
and p-stacking[33] have been discussed, mainly in the context
of selectivity.
ˆ
51.7 (CH3), 67.8 (CH2), 169.9 (C O); HRMS calcd for C5H6O4 [M À
MeOH] 130.0266609, found 130.026352.
Typical procedure for the enzymatic reaction: (Æ)-trans-Cyclohexane-1,2-
diamine ((Æ)-3, 0.60 mL, 5.0 mmol), dimethyl 3-oxapentanedioate (4,
0.86 g, 5.0 mmol), and 1,4-dioxane (20 mL) were added to an Erlenmeyer
flask containing lipase B from Candida antarctica (0.30 g) under nitrogen.
The reaction mixture was shaken at 308C and 250 rpm until no traces of the
diester were detected by TLC (acetone:diethyl ether:propan-2-ol
10:20:0.2). Afterwards, the enzyme was separated by filtration and washed
with dichloromethane. Dry HCl was bubbled through the organic solution
until a white precipitate was formed. Then the solvents were evaporated to
dryness and the residue extracted with dichloromethane and 3n HCl. The
organic layer was dried and evaporated, yielding the bis(amidoester)
(R,R)-5. The aqueous layers contained a mixture of the monoamidoester
and the diamine (S,S)-3 as their hydrochloride salts, which can be separated
by precipitation of (S,S)-3 ´ 2HCl with ethanol.
(R,R)-N,N'-(Cyclohexane-1,2-diyl)bis(methyl
4-carbamoyl-3-oxabuta-
noate) [(R,R)-5]: White solid (560 mg, 30% yield). M.p. 82± 84 8C;
In summary, we have described a simple and efficient
chemoenzymatic method for the synthesis of two new 18-
membered-ring oxaazamacrocycles from racemic trans-cyclo-
hexane-1,2-diamine. Their protonation behavior and their
binding constants with different chiral anions have been
studied. These compounds show enantiomeric and diastereo-
meric molecular recognition properties depending on the
structure of the anion added. From the experimental results
we propose a model for the complexation based on two
coulombic and additional hydrogen-bond interactions. It must
[a]2D0 ˆ ‡38.3 (c ˆ 0.65 in CHCl3); ee >99%; IR (nujol): 3295 (NH), 1754
(C O), 1640 (C O) cmÀ1; 1H NMR (CDCl3, 200 MHz): d ˆ 1.35 (brd, J ˆ
4.9 Hz, 4H; 2CH2), 1.78 (brs, 2H; CH2), 2.05 (brs, 2H; CH2), 3.75 (s, 8H;
2CH ‡ 2CH3), 3.92± 4.14 (AB q, JAB ˆ 15.1 Hz, dA ˆ 3.97, dB ˆ 4.09, 4H;
2CH2), 7.08 (brs, 2H; NH); 13C NMR (CDCl3, 50.3 MHz): d ˆ 24.9 (CH2),
ˆ
ˆ
ˆ
32.5 (CH2), 52.4 (CH), 53.1 (CH3), 68.9 (CH2), 71.3 (CH2), 169.6 (C O),
170.6 (C O); MS (10 eV, E.I.) m/z (%): 374 (<5) [M] , 227 (100);
‡
ˆ
C16H26N2O8 (374.4): calcd C 51.33, H 7.00, N 7.48; found C 51.46, H 6.99, N
7.36; HPLC conditions for the determination of the enantiomeric excess:
Chiralcel OD column, isocratic conditions, hexane:ethanol 90:10, sample
conc: 0.5 mgmLÀ1, flow rate: 0.8 mLminÀ1, retention times (min): 15.7
(R,R), 19.8 (S,S), RS ˆ 2.75.
‡
be pointed out that, as far as we know, (R,R)-1 ´ 4H seems to
(R,R)-N,N'-(Cyclohexane-1,2-diyl)bis(4-carbamoyl-3-oxabutanamide)
[(R,R)-6]: Bis(amidoester) (R,R)-5 (1.87 g, 5.0 mmol) was added to a
solution of NH3 in methanol (10%, 150 mL) and stirred at 78C over 12h.
Then the solvent was removed and (R,R)-6 was obtained quantitatively as a
be the best receptor for the recognition of an anion with a
chiral center in aqueous solution.
white solid. M.p. 96 ± 978C; [a]D20 ˆ ‡16.1 (c ˆ 0.28 in MeOH); IR (nujol)
À1
ˆ
ˆ
3505 (NH), 3295 (NH), 3151 (NH), 1682 (C O), 1646 (C O) cm
;
1H NMR (CDCl3, 200 MHz): d ˆ 1.35 (brs, 4H; 2CH2), 1.83 (brs, 2H;
Experimental Section
CH2), 2.1 (brs, 2H; CH2), 3.78 (brs, 2H; 2CH), 3.92± 4.08 (AB q, JAB
ˆ
General: All reagents were purchased from Aldrich Chemie. Solvents were
distilled over an adequate desiccant and stored under nitrogen. Flash
chromatography was performed with Merck silica gel 60 (230 ± 240).
Melting points were taken on a Gallenkamp apparatus and are uncorrect-
ed. Optical rotations were measured by means of a Perkin ± Elmer 241
polarimeter. IR spectra were recorded on a Perkin ± Elmer 1720-X FT IR
spectrometer. Mass spectra were recorded on a VG Autospec. Micro-
14.7 Hz, dA ˆ 3.97, dB ˆ 4.02, 4H; 2CH2), 3.95 ± 4.16 (AB q, JAB ˆ 15.3 Hz,
dA ˆ 4.00, dB ˆ 4.11, 4H; 2CH2), 5.65 (brs, 2H; NH), 7.00 (brs, 2H; NH),
7.10 (brs, 2H; NH); 13C NMR ([D6]DMSO, 50.3 MHz): d ˆ 24.5 (CH2), 31.5
‡
(CH2), 52.2 (CH), 70.0 (2CH2), 169.0 (C), 171.2(C); MS (FAB , nitrobenzyl
‡
‡
alcohol matrix) m/z (%): 345 (76) [M‡H] , 367 (100) [M‡Na] ;
C14H24N4O6 (344.4): calcd C 48.83, H 7.03, N 16.27; found C 48.53, H
6.99, N 16.36.
3336
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3336 $ 17.50+.50/0
Chem.Eur.J. 2000, 6, No. 18

Dioxatetraazamacrocycles
3331±3338
(R,R)-N,N'-(Cyclohexane-1,2-diyl)bis(3-oxa-pentane-1,5-diamine) tetra-
hydrochloride [(R,R)-7 ´ 4HCl]: A solution of 1m BH3 ´ THF (100 mL)
was added dropwise to a suspension of (R,R)-6 (1.72g, 5.0 mmol) in dry
tetrahydrofuran (200 mL) under nitrogen. The mixture was refluxed for
7 h. After this time, the reaction mixture was treated at 08C with water
(20 mL) and evaporated to dryness. The solid thus obtained was refluxed in
6n HCl (200 mL) for 4 h and then the water was removed in vacuo. The
residue was dissolved in the minimum amount of H2O, passed over Dowex
1 (basic form), and eluted with deionized water. The aqueous solution
containing (R,R)-7 was acidified to pH ˆ 2with conc. HCl and the solvent
evaporated, and (R,R)-7 ´ 4HCl (2.08 g, 96%) was obtained as a highly
hygroscopic solid after recrystallization in MeOH/CHCl3. [a]2D0 ˆ À32.0
(c ˆ 1.0 in MeOH); 1H NMR (D2O, 300 MHz): d ˆ 1.32(brm, 2H; 2 HCH),
1.52(brs, 2H; 2 HCH), 1.73 (brs, 2H; 2HCH), 2.24 (brd, J ˆ 13.1 Hz, 2H;
2HCH), 3.16 (t, J ˆ 5.0 Hz, 4H; 2CH2), 3.24 (m, 2H; 2CH), 3.47 (m, 4H;
2CH2), 3.71 (t, J ˆ 4.6 Hz, 4H; 2CH2), 3.79 (t, J ˆ 4.2Hz, 4H; 2CH 2); MS
80%; white solid; m.p. 200 ± 2028C; [a]2D0 ˆ ‡5.2( c ˆ 0.65 in CHCl3);
1H NMR (CDCl3, 200 MHz): d ˆ 0.95 ± 1.40 (several m, 8H; 4CH2), 1.50 ±
1.75 (several m, 8H; 4CH2), 2.43 (s, 12H; 4CH3), 2.85 ± 4.10 (several m,
18H; 8CH2 ‡ 2CH), 4.91 (brs, 2H; 2CH), 7.31 (d, J ˆ 7.9 Hz, 8H; 8CH
‡
arom), 7.84 (brs, 4H; 4CH arom), 8.02(brs, 4H; 4CH arom); MS (FAB
,
‡
‡
nitrobenzyl alcohol matrix) m/z (%): 985 (3) [M‡H] , 1007 (10) [M‡Na] ,
‡
775 (5) [M À Ts] ; C48H64N4O10S4 (985.3): calcd C 58.51, H 6.55, N 5.69;
found C 58.64, H 6.25, N 5.74.
General procedure for the hydrolysis of the tetratosylated azamacrocycles
9 and 11: Tetratosylated azamacrocycle (0.70 mmol) and an excess of
phenol (1.14 mL, 13 mmol) were dissolved in 48% aqueous HBr (21 mL),
and the solution was heated under reflux for 4 days. Water and dichloro-
methane were then added and the aqueous layer was repeatedly washed
with dichloromethane. The organic layer was discarded and the aqueous
one was evaporated under reduced pressure. The residue was treated with
4n NaOH (5 mL) and extracted several times with dichloromethane. The
combined organic layers were dried and evaporated to dryness, yielding the
desired azamacrocycle as a colorless oil. For the potentiometric measure-
ments and for storage, the tetrahydrochloride was prepared by dissolution
of the free amine in methanol, addition of conc. HCl and evaporation of the
solvents. The residue was recrystallized in EtOH to yield the tetrahydro-
chloride of the corresponding dioxatetraazamacrocycle as a highly hygro-
scopic white solid, which is dried in vacuo and kept in a desiccant.
‡
‡
(FAB , nitrobenzyl alcohol matrix) m/z (%): 289 (100) [M‡1] , 311 (10)
‡
[M‡Na] ; C16H36N4Cl4 (458.4): calcd C 38.72, H 8.36, N 12.90; found C
38.46, H 8.12, N 13.10. The unprotonated species may be obtained by
addition of sodium hydroxide to the aqueous solution of the corresponding
tetrahydrochloride to pH ˆ 13. Extraction with dichloromethane and
subsequent evaporation of the solvent yielded (R,R)-7 as an oily substance
unsuitable for storage; 1H NMR (CDCl3, 200 MHz): d ˆ 1.05 (m, 4H,
4NH), 1.10 (t, J ˆ 7.1 Hz, 2H; 2NH), 1.24 (brt, J ˆ 9.0 Hz, 2H; 2HCH), 1.73
(brd, J ˆ 8.2Hz, 2H; 2 HCH), 2.04 (brs, 2H; 2CH), 2.14 (brd, J ˆ 11.4 Hz,
2H; 2HCH), 2.58 ± 2.69 (m, 2H; 2HCH), 2.85 (t, J ˆ 5.2Hz, 4H; 2CH 2),
2.87 ± 2.98 (m, 2H; 2HCH), 3.48 (t, J ˆ 5.2Hz, 4H; 2CH 2), 3.52± 3.60 (m,
4H; 2HCH), 4.80 (brs, 2H, NH); 13C NMR (CDCl3, 50.3 MHz): d ˆ 25.3
(CH2), 31.8 (CH2), 41.9 (CH2), 46.6 (CH2), 61.8 (CH), 71.0 (CH2), 73.1
(CH2).
(1R,18R)-5,14-Dioxa-2,8,11,17-tetraazabicyclo[16.4.0]docosane tetrahy-
drochloride [(R,R)-1 ´ 4HCl]. Yield: 88%; decomposes at 2508C; [a]D20
ˆ
À35.6 (c ˆ 0.50 in H2O); 1H NMR (D2O, 300 MHz): d ˆ 1.10 ± 1.55 (m, 4H;
2CH2), 1.65 (m, 2H; 2HCH), 2.05 ± 2.45 (m, 2H; 2HCH), 3.10 ± 3.90 (m,
‡
22H; 10CH2 ‡ 2CH); MS (FAB , nitrobenzyl alcohol matrix) m/z (%): 315
‡
(39) [M‡1] ; C16H38N4O2Cl4 (460.4): calcd C 41.74, H 8.32, N 12.17; found
C 41.42, H 8.38, N 12.41. Spectral data for the free amine (R,R)-1: 1H NMR
(DCCl3, 300 MHz): d ˆ 0.90 ± 1.35 (m, 6H; 2CH2 ‡ 2HCH), 1.72(brd, J ˆ
11.3 Hz, 2H; 2HCH), 2.10 (brd, J ˆ 10.9 Hz, 2H; 2HCH), 2.24 (m, 2H;
2CH), 2.30 ± 2.50 (b, 4H, 4NH), 2.61 (ddd, J ˆ 12.6, 6.54, and 2.62 Hz, 2H;
2HCH), 2.79 (s, 4H; 2CH2), 2.80 (m, 4H; 2CH2), 2.97 (ddd, J ˆ 12.6, 6.54,
and 2.62 Hz, 2H; 2HCH), 3.48 ± 3.66 (m, 8H; 4CH2); 13C NMR (DCCl3,
100.7 MHz): d ˆ 23.1 (CH2), 26.9 (CH2), 44.9 (CH2), 46.0 (CH2), 48.4 (CH2),
58.7 (CH), 65.9 (CH2), 66.0 (CH2).
(R,R)-N,N'-(Cyclohexane-1,2-diyl)-N,N',N'',N'''-tetrakis(p-toluenesulfon-
yl)bis(3-oxapentane-1,5-diamine) [(R,R)-8]: Tetrahydrochloride (R,R)-7 ´
4HCl (700 mg, 1.61 mmol), diethyl ether (13 mL), and p-toluenesulfonyl
chloride (1.71 g, 9.02mmol) were added to a suspension of K 2CO3 (12.5 g)
in water (36 mL). The resulting biphasic system was vigorously stirred for
12h. Water and ethyl acetate were then added to the mixture. The organic
layer was washed with 3n HCl, dried, and evaporated to give (R,R)-8 in
76% yield after flash chromatography (ethyl acetate/hexane 2:3). M.p. 86 ±
(1S,9S,14S,22S)-5,18-Dioxa-2,8,15,21-tetraazatricyclo[20.4.0.09,14]hexaco-
sane tetrahydrochloride [(S,S,S,S)-2 ´ 4HCl]: Yield: 75%; decomposes at
2458C; [a]2D0 ˆ À33.2( c ˆ 0.50 in H2O); 1H NMR (D2O, 200 MHz): d ˆ 1.28
(m, 4H; 4HCH), 1.58 (brm, 4H; 4HCH), 1.70 (m, 4H), 2.07 (brd, J ˆ
12.9 Hz, 4H; 4HCH), 3.20 (m, 4H; 4HCH), 3.40 (m, 4H; 4HCH), 3.64 (m,
4H; 4CH), 3.85 (brd, J ˆ 6.46 Hz, 8H; 4CH2); 13C NMR (D2O, 75.5 MHz):
1
888C; [a]2D0 ˆ ‡6.7 (c ˆ 0.51 in CHCl3); H NMR (CDCl3, 200 MHz): d ˆ
1.10 ± 1.40 (several m, 8H; 4CH2), 1.60 (brs, 2H; 2HCH), 2.38 (s, 6H;
2CH3), 2.46 (s, 6H; 2CH3), 3.09 ± 3.20 (m, 8H; 4CH2), 3.38 ± 3.48 (m, 2H;
2HCH), 3.52± 3.62 (m, 2H, 2 HCH), 3.78 (t, J ˆ 7.3 Hz, 4H; 2CH2), 3.90
(brs, 2H; 2CH), 5.37 (t, J ˆ 5.8 Hz, 2H, NH), 7.20 (d, J ˆ 8.2Hz, 4H; 4CH
arom), 7.35 (d, J ˆ 8.2Hz, 4H; 4CH arom), 7.71 (d, J ˆ 8.2Hz, 4H; 4CH
arom), 7.74 (d, J ˆ 8.2Hz, 4H; 4CH arom); 13C NMR (CDCl3, 50.3 MHz):
d ˆ 21.4 (CH3), 21.5 (CH3), 24.9 (CH2), 29.6 (CH2), 42.5 (CH2), 42.8 (CH2),
59.4 (CH), 69.1 (CH2), 69.8 (CH2), 127.0 (CH), 127.4 (CH), 129.6 (CH),
‡
d ˆ 23.5 (CH2), 28.9 (CH2), 42.1 (CH2), 56.2(CH), 63.3 (CH 2); MS (ES ,
‡
‡
2 5 V)m/z (rel. intensity): 369.2(2) [ M‡1] , 387.2(2) [ M‡1‡H2O] , 405.2
(2.6) [M‡1‡H35Cl] , 407.2(1.3) [ M‡1‡H37Cl] ; C20H44N4O2Cl4 (514.5):
‡
‡
calcd C 46.69, H 8.62, N 10.89; found C 46.42, H 8.38, N 11.01.
‡
129.8 (CH), 136.7 (C), 137.0 (C), 143.1 (C), 143.8 (C); MS (FAB ,
‡
nitrobenzyl alcohol matrix) m/z (%): 905 (43) [M‡H] , 927 (100)
‡
‡
[M‡Na] , 749 (61) [M À Ts] ; C42H56N4O10S4 (905.2): calcd C 55.73, H
6.24, N 6.19; found C 55.59, H 6.60, N 5.81.
Acknowledgements
(1R,18R)-2,8,11,17-Tetrakis(p-toluenesulfonyl)-5,14-dioxa-2,8,11,17-tetra-
azabicyclo[16.4.0]docosane [(R,R)-9]: Dry acetonitrile (25 mL) was added
to a flask containing (R,R)-8 (765 mg, 0.85 mmol) and Cs2CO3 (2.76 g,
8.46 mmol) under nitrogen. The mixture was refluxed for half an hour and
then a solution of ethylene glycol dimesylate (184 mg, 0.85 mmol) in dry
acetonitrile (17 mL) was added dropwise. The reaction mixture was kept
under reflux for 4 days. After this time, the solvent was removed and the
residue subjected to flash chromatography (ethyl acetate:hexane 2:3) to
yield (R,R)-9 (87%) as a white solid. M.p. 198 ± 2008C; [a]2D0 ˆ À11.2( c ˆ
0.50 in CHCl3); 1H NMR (CDCl3, 200 MHz): d ˆ 1.00 ± 1.50 (brm, 4H;
2CH2), 1.50 ± 1.80 (brm, 4H; 2CH2), 2.44 (m, 12H; 4CH3), 3.10 ± 4.00
(several m, 22H; 10CH2 ‡ 2CH), 7.31 (m, 8H; 8CH arom), 7.73 (m, 8H;
This work was supported by the CICYT (BIO98-0770). I.A. thanks the
Â
Ministerio de Educacion y Ciencia for a predoctoral fellowship.
[1] a) J. M. Lehn, Supramolecular Chemistry, VCH, Weinheim, 1995;
b) J. M. Lehn, Angew.Chem. 1988, 100, 91-116; Angew.Chem.Int.Ed.
Engl. 1988, 27, 89 ± 112; c) J. M. Lehn, Angew.Chem. 1990, 102, 1347 ±
1362; Angew.Chem.Int.Ed.Engl . 1990, 29, 1304 ± 1319.
[2] R. M. Izatt, K. Pawlak, S. S. Bradshaw, R. L. Bruening, Chem.Rev .
1995, 95, 2529 ± 2586.
[3] R. D. Hancock, R. Bhavan, P. W. Wade, J. C. A. Boeyens, S. M.
Dobson, Inorg.Chem. 1989, 28, 187 ± 194.
[4] a) F. P. Schmidtchen, M. Berger, Chem.Rev. 1997, 97, 1609 ± 1646;
b) R. J. Motekaitis, W. B. Utley, A. E. Martell, Inorg.Chim.Acta 1993,
212, 15 ± 21; c) R. J. Motekaitis, A. E. Martell, Inorg.Chem. 1992, 31,
5534 ± 5542.
‡
8CH arom); MS (FAB , nitrobenzyl alcohol matrix) m/z (%): 931 (28)
‡
‡
‡
[M‡H] , 953 (96) [M‡Na] , 775 (39) [M À Ts] ; C44H58N4O10S4 (931.2):
calcd C 56.75, H 6.28, N 6.02; found C 56.54, H 6.24, N 5.74.
(S,S)-Cyclohexane-1,2-diylbis(sulfonamide) [(S,S)-10] was prepared from
(S,S)-3 ´ 2HCl as previously described. [22]
(1S,9S,14S,22S)-2,8,15,21-Tetrakis(p-toluenesulfonyl)-5,18-dioxa-2,8,15,21-
tetraazatricyclo[20.4.0.09,14]hexacosane [(S,S,S,S)-11] was obtained from
(S,S)-10 and diethylene glycol dimesylate as described for (R,R)-9. Yield:
[5] a) M. W. Hosseini, J. M. Lehn, M. P. Mertes, Helv.Chim.Acta 1983,
66, 2454 ± 2466; b) M. W. Hosseini, J. M. Lehn, L. Maggiora, K. B.
Chem.Eur.J. 2000, 6, No. 18
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3337 $ 17.50+.50/0
3337

V. Gotor et al.
FULL PAPER
Mertes, M. P. Mertes, J.Am.Chem.Soc. 1987, 109, 537 ± 544; c) Q. Lu,
A. E. Martell, R. J. Motekaitis, Inorg.Chim.Acta 1996, 251, 365 ± 370.
[6] a) M. W. Hosseini, J. M. Lehn, Chem.Commun. 1988, 397 ± 399;
b) M. P. Mertes, K. B. Mertes, Acc.Chem.Res. 1990, 23, 413 ± 418.
[7] H. Fenniri, J. M. Lehn, A. Marquis-Rigault, Angew.Chem . 1996, 108,
337 ± 339; Angew.Chem.Int.Ed.Engl . 1996, 35, 337 ± 339.
[8] K. Kano, H. Kamo, S. Negi, T. Kitae, R. Takaoka, M. Yamaguchi, H.
Okubo, M. Hirama, J.Chem.Soc.Perkin Trans.2 1999, 15 ± 2 1.
[9] T. Kitae, H. Takashima, K. Kano, J.Incl.Phenom. 1999, 33, 345 ± 359.
[10] J. L. Sessler, A. Andrievsky, V. Kral, V. Lynch, J.Am.Chem.Soc. 1997,
119, 9385 ± 9392.
[21] F. A. Carey, R. J. Sundberg, Advanced Organic Chemistry.Part A:
Structure and Mechanisms, Plenum, New York, 1993.
[22] I. Alfonso, C. Astorga, F. Rebolledo, V. Gotor, Tetrahedron:
Asymmetry 1999, 10, 2515 ± 2522.
[23] B. Dietrich, M. W. Hosseini, J. M. Lehn, R. B. Sessions, Helv.Chim.
Acta 1983, 1262 ± 1278.
[24] R. Barbucci, P. Paoletti, A. Vacca, J.Chem.Soc.A 1970, 2202 ± 2206.
[25] C. Bertsc, W. C. Fernelius, P. B. Block, Pennsylvania State Univ.
Contract No. AT(30-1)-907, April 30, 1956.
[26] D. Voet, J. G. Voet, Biochemistry, 2nd ed., Wiley, New York, 1995.
[27] a) G. J. Siegel, B. W. Agranoff, R. W. Albers, P. B. Molinoff, Basic
Neurochemistry: Molecular, Cellular and Medical Aspects, 5th ed.,
Raven, New York, 1994, pp. 367 ± 387; b) J. I. Yeh, H. P. Biemann,
G. G. Prive, J. Pandit, D. E. Koshland, Jr., S. H. Kim, J.Mol.Biol.
1996, 262, 186 ± 201.
[28] T. H. Webb, C. S. Wilcox, Chem.Soc.Rev. 1993, 383 ± 395.
[29] G. Papogan, K. Gu, J. Wiorkiewicz-Kuczera, K. Kuczera, K. Bowman-
James, J.Am.Chem.Soc. 1996, 118, 1354 ± 1364.
[11] K. E. Krakowiak, J. S. Bradshaw, D. J. Zamecka-Krakowiak, Chem.
Rev. 1989, 89, 929 ± 972.
[12] D. J. Cram, J. Cram, Acc.Chem.Res. 1978, 11, 8 ± 14.
[13] a) A. Bianchi, K. Bowman-James, E. García-EspanÄa, The Supra-
molecular Chemistry of Anions, VCH, Weinheim, 1997; b) B. Dietrich,
Pure Appl.Chem. 1993, 65, 1457 ± 1464; c) C. Seel, J. De Mendoza, in
Comprehensive Supramolecular Chemistry, Pergamon, Oxford, 1996,
pp. 519 ± 552.
[30] MMX force field as implemented in PCMODEL V5.0, Serena
Software, Blomington. All the calculations were performed with the
dielectric constant of water.
[14] For some examples see: a) B. M. Trost, R. C. Bunt, Angew.Chem .
1996, 108, 70 ± 73; Angew.Chem.Int.Ed.Engl
. 1996, 35, 99 ± 102;
b) B. M. Trost, R. Radinov, E. M. Grenzer, J.Am.Chem.Soc. 1997,
119, 7879 ± 7880; c) M. Palucki, N. S. Finney, P. J. Pospisil, M. L. Güler,
[31] a) C. Bazzicalupi, A. Bencini, A. Bianchi, M. Cecchi, B. Escuder, V.
Fusi, E. García-EspanÄa, C. Giorgi, S. V. Luis, G. Maccagni, V.
T. Ishida, E. N. Jacobsen, J.Am.Chem.Soc.
d) S. E. Schaus, J. Branalt, E. N. Jacobsen, J.Org.Chem. 1998, 63,
403 ± 405.
1998, 120, 948 ± 954;
Marcelino, P. Paoletti, B. Valtancoli, J.Am.Chem.Soc.
6807 ± 6815; b) M. Berger, F. P. Schmidtchen, Angew.Chem. 1998, 110,
2840 ± 2842; Angew.Chem.Int.Ed.Engl. 1998, 37, 2694 ± 2696; c) B.
1999, 121,
[15] a) N. Farrell, D. M. Kiley, W. Schmidt, M. P. Hacker, Inorg.Chem .
1990, 29, 397 ± 403; b) K. Skog, O. Wennerström, Tetrahedron Lett.
1992, 33, 1751 ± 1754; c) S. S. Yoon, W. C. Still, J.Am.Chem.Soc. 1993,
115, 823 ± 824; d) W. C. Still, Acc.Chem.Res. 1996, 29, 155 ± 163; e) J.
Reedijk, Chem.Commun. 1996, 801 ± 806; f) Z. Pan, W. C. Still,
Tetrahedron Lett. 1996, 37, 8699 ± 8702; g) F. Gasparrini, D. Misiti,
W. C. Still, C. Villani, H. Wennemers, J.Org.Chem. 1997, 62, 82 2 1 ±
8224.
Linton, A. D. Hamilton, Tetrahedron 1999, 55, 6027 ± 6038.
[32] Hydrogen bonds are of considerable importance in anion ± macrocy-
cle interaction, and enthalpic ± entropic compensatory relationships
have been proposed for such complexation reactions (ref. [31]a).
[33] a) M. I. Burguete, P. Díaz, E. García-EspanÄa, S. V. Luis, J. F. Miravet,
M. Querol, J. A. Ramírez, Chem.Commun. 1999, 649 ± 650; b) C.
Bazzicalupi, A. Beconcini, V. Fusi, C. Giorgi, A. Masotti, B.
Valtancoli, J.Chem.Soc.Perkin Trans.2
1999, 1675 ± 682; c) M. T.
[16] E. Kimura, A. Sakonaka, T. Yatsunami, M. Kodama, J.Am.Chem.
Soc. 1981, 103, 3041 ± 3045.
[17] For other sequential kinetic resolutions of diols see: a) R. J. Kazlaus-
kas, J.Am.Chem.Soc. 1989, 111, 4953 ± 4959; b) Z. W. Guo, S. H. Wu,
C. S. Chen, G. Girdaukas and C. J. Sih, J.Am.Chem.Soc. 1990, 112,
4942± 4945.
Albelda, M. A. Bernardo, E. García-EspanÄa, M. L. Godino-Salido,
S. V. Luis, M. J. Melo, F. Pina, C. Soriano, J.Chem.Soc.Perkin Trans.2
1999, 1675 ± 682; d) T. Paris, J. P. Vigneron, J. M. Lehn, M. Cesario, J.
Guilhen, C. Pascard, J.Incl.Phenom.
1999, 33, 191 ± 202; e) O.
Baudoin, F. Gonnet, M. P. Teulade-Fichou, J. P. Vigneron, J. C. Tabet,
J. M. Lehn, Chem.Eur.J. 1999, 5, 2762 ± 2771.
[18] I. Alfonso, C. Astorga, F. Rebolledo, V. Gotor, Chem.Commun. 1996,
2471 ± 2472.
[34] P. Gans, A. Sabatini, A. Vacca, J.Chem.Soc.Dalton Trans.
1985,
1195 ± 1200.
[19] a) J. E. Richman, T. J. Atkins, J.Am.Chem.Soc . 1974, 96, 2268 ± 2270;
b) B. K. Vriesema, J. Butler, R. M. Kellog, J.Org.Chem . 1984, 49,
110 ± 113.
Received: August 4, 1999
[20] J. F. Biernat, E. Luboch, Tetrahedron 1984, 40, 1927 ± 1929.
Revised version: March 13, 2000 [F1958]
3338
ꢀ WILEY-VCH Verlag GmbH, D-69451 Weinheim, 2000
0947-6539/00/0618-3338 $ 17.50+.50/0
Chem.Eur.J. 2000, 6, No. 18