Steric and stereoelectronic effects in aza crown ether complexes

Article abstract of DOI:10.1002/(SICI)1099-0690(199807)1998:7<1379::AID-EJOC1379>3.0.CO;2-9

Stability constants and enthalpy changes determined by calorimetric titrations and supported by selected NMR titrations are reported for the complexation of sodium and potassium cations with 18 different crown ethers containing nitrogen atoms with different number, location and substitution pattern. The data, measured in methanol mostly with potassium salts, are compared to literature data; they show striking differences between all-oxygen analogs and the macrocycles with NH groups. In contrast, affinities with aza crown ethers bearing alkyl groups at the nitrogen as well as with the cryptand [2.2.2] come closer to the complexation free energies predicted from the number and electron donating capacity of the ligand heteroatoms. This is rationalised on the basis of molecular mechanics calculations, showing that a NH-containing crown predominates in conformations with axial N lone pairs, due to their repulsive electrostatic interactions with the ring oxygen atoms. Replacement of the hydrogen by alkyl groups forces the lone pairs to an equatorial position, thus enabling better complex formation, as borne out by experiment. In line with these arguments the IgK differences are with some exceptions more due to ΔH than to TΔS differences. The calorimetric data show linear isoequilibrium correlations between TΔS and ΔH, with slopes between those observed with other crown ether and cryptand complexes. Preliminary investigations of some synthetic macrocyclic amide precursors yield appreciable complexation only, if the two carbonyl oxygens can come in close contact with the guest cation. Computer aided molecular modelling shows that this is possible in a small 15C5-derivative, in which the polyethylenglycol cycle only serves as ring template without binding contributions from the ether oxygen atoms.

Full text of DOI:10.1002/(SICI)1099-0690(199807)1998:7<1379::AID-EJOC1379>3.0.CO;2-9

FULL PAPER
Steric and Stereoelectronic Effects in Aza Crown Ether Complexes[1]
Vitally P. Solov’eva, Nadezhda N. Strakhovaa, Vladimir P. Kazachenkoa, Alexandr F. Solotnova, Vladimir E. Baulina,
Oleg A. Raevskya, Volker Rüdigerb, Frank Eblingerb, and Hans-Jörg Schneider*b
Institute of Physiologically Active Compounds, Russian Academy of Sciencea,
142432, Chernogolovka, Moscow Region, Russia
FR Organische Chemie, Universität des Saarlandesb,
D-66041 Saarbrücken, Germany
Received December 22, 1997
Keywords: Calorimetry / Molecular modelling / Crown compounds / Cryptands / Macrocyclic ligands / Conformational
analysis
Stability constants and enthalpy changes determined by
to their repulsive electrostatic interactions with the ring
calorimetric titrations and supported by selected NMR oxygen atoms. Replacement of the hydrogen by alkyl groups
titrations are reported for the complexation of sodium and
potassium cations with 18 different crown ethers containing
nitrogen atoms with different number, location and
substitution pattern. The data, measured in methanol mostly
forces the lone pairs to an equatorial position, thus enabling
better complex formation, as borne out by experiment. In line
with these arguments the lgK differences are with some
exceptions more due to ∆H than to T∆S differences. The
with potassium salts, are compared to literature data; they calorimetric data show linear isoequilibrium correlations
show striking differences between all-oxygen analogs and
the macrocycles with NH groups. In contrast, affinities with
aza crown ethers bearing alkyl groups at the nitrogen as well
between T∆S and ∆H, with slopes between those observed
with other crown ether and cryptand complexes. Preliminary
investigations of some synthetic macrocyclic amide
as with the cryptand [2.2.2] come closer to the complexation precursors yield appreciable complexation only, if the two
free energies predicted from the number and electron
donating capacity of the ligand heteroatoms. This is
carbonyl oxygens can come in close contact with the guest
cation. Computer aided molecular modelling shows that this
rationalised on the basis of molecular mechanics is possible in
a small 15C5-derivative, in which the
calculations, showing that
predominates in conformations with axial N lone pairs, due
a
NH-containing crown
polyethylenglycol cycle only serves as ring template without
binding contributions from the ether oxygen atoms.
In spite of countless investigations on synthetic ionophor with some aza crown ether complexes. Thus, the stabilities
complexes[2] the underlying binding mechanisms and the of potassium complexes of the cryptand [2.2.2] (in meth-
predictability of complex stabilities require further studies. anol ∆G ϭ 60 kJ/mol, see Table 1) are quite in line with the
One aim of our work is the prediction of crown ether and ∆G ϭ 57 kJ/mol which we predict from the number and
cryptand stabilities by as simple rules as possible, based on electron donor capacity of the ligand heteroatoms. How-
additive interaction increments between the ligand het- ever, the underlying 1,10-diaza-crownether, which has only
eroatoms and the metal ion.[3] We have shown, that as long 2 out of 8 heteroatoms less than the cryptand, shows with
as strainfree matching between ligand sites and ion is pos- ∆G ϭ 10 kJ/mol an affinity which is very much smaller,
sible dozens of complexation constants can be predicted also in comparison to that expected on the basis of simple
with a standard deviation of ∆lgK ϭ 0.25, if the electron additive interactions. In order to obtain insight into the ori-
donor capacity sum of the different coordination centres Cd gin of these apparent discrepancies, and to find a lead for
of ligand is taken from measurements of hydrogen bond predicting conditions for optimal complexation we studied
complexes of simple compounds with corresponding elec- a number of aza crown ether complexes in which substi-
tron donor centres in lipophilic solvents.[3][4] Medium ef- tution patterns and position of the nitrogen atoms was va-
fects on crown ether and cryptand complex stabilities in ried (Scheme 1). Furthermore we wanted to analyse en-
various solvents can be described essentially by the change thalpic vs. entropic binding differences by calorimetric ti-
of free solvation energy of the ion in these solvents.[5] How trations with several aza crown ether complexes. Until now
mismatch between ligand sites and the cation lowers the there are only scattered reports on the thermodynamics of
affinities has been analysed on the basis of computer aided azacrownether complexes[7][8] most frequently with tran-
molecular modelling and NMR measurements with 18C5- sition metal cations. Table 1 also shows literature values[9]
potassium complexes.[6]
which are relevant in the context of the present work, often
The present paper mainly addresses substantial, and at showing large deviations. These might be due partially to
first sight unexpected variations in binding energies ∆G the pH dependence as consequence of partial nitrogen pro-
Eur. J. Org. Chem. 1998, 1379Ϫ1389
WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0707Ϫ1379 $ 17.50ϩ.50/0
1379

H.-J. Schneider
FULL PAPER
tonation and H-bonding, and/or to neglected anion effects, elling with the aid of the CHARMm field[10] helps to eluci-
and made it necessary to measure these complexes again date the origin of the unusual instability of the Kϩ-2a
˚
under the same conditions as the new ones.
complex. Molecular dynamics simulations in a 15A di-
Table 1. Stability constants (lgβ units) and thermodynamic parameters (kJ/mol) for complexation of azacrown ethers with potassium
cation in methanol at 298 K.[a]
No
Ligand
18C6
Salt
M:L[b]
lgβ[c]
Ϫ∆G
Ϫ∆H
T∆S
1
KSCN
KI
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:2
1:1
1:1
1:1
1:1
1:1
1:1
1:2
1:1
1:2
1:1
1:2
1:1
1:1
1:2
1:1
>5.7
>32.5
54.7(1.0)
>Ϫ-22.2
6.22(0.04)[d]
35.5(0.2)[d]
Kϩ
6.07
34.6[9a]
56.8[9a]
3.2(0.3)
Ϫ22.2[9a]
9.6(1.4)
[9a]
2a
1,10-N218C6
KOH
KI
2.25(0.25)
1.76(0.03)[d]
1.83[9b]
12.8(1.4)
10.0(0.2)[d]
10.4[9b]
Kϩ
4.7[9b]
5.7[9b]
Kϩ
2.04[9h]
11.6[9h]
Kϩ
11.2(1)[9k]
25.7(1.1)
24.8(0.7)
29.6(0.3)[d]
23.4[9c]
2b
1,10-(MeN)218C6
KCl
KOH
KI
4.51(0.20)
4.35(0.13)
5.18(0.06)[d]
4.10[9c]
33.6(1.5)
Ϫ7.8(1.9)
Kϩ
Kϩ
5.3[9d]
30.2[9d]
2c
3a
1-BzN-10-N18C6
N18C6
KCl
3.72(0.05)
6.16(0.08)
3.78(0.19)
3.67(0.08)
4.18[9i]
21.2(0.3)
35.2(0.4)
21.6(1.1)
20.9(0.4)
23.9[9i]
14.1(0.1)
26.0(0.6)
22.0(1.2)
20.7(1.0)
7.1(0.3)
9.2(0.7)
Ϫ0.4(1.6)
0.2(1.1)
KNO3
KOH
Kϩ
Kϩ
3.90[9h]
22.3[9h]
3b
3c
MeN18C6
BzN18C6
KI
5.46(0.07)[d]
4.83(0.66)
7.78(1.08)
4.17(0.42)
5.97(0.44)
4.92(0.58)
8.60(0.73)
2.1(0.2)
31.2(0.4)[d]
27.6(3.8)
44.4(6.2)
23.8(2.4)
34.1(2.5)
28.1(3.3)
49.1(4.2)
12(1)
KNO3
44.5(2.2)
64.2(5.6)
1.4(0.1)
6.1(0.6)
28.5(1.8)
44.6(1.3)
Ϫ2.2(0.5)
13(6)
Ϫ16.9(4.4)
Ϫ19.8(8.4)
22.4(2.4)
28.0(2.6)
Ϫ0.4(3.8)
4.5(4.4)
14(1)
2(6)
20(1)
7.1(3.2)
11.1(3.1)
4a
4b
1,7-N218C6
KCl
KCl
1,7-(MeN)218C6
4c
5a
8,18-Dioxo-1,7-N218C6
1,4-N218C6
KCl
KCl
2.6(0.4)
14.8(2.3)
31.4(1.1)
26.9(3.1)
48.2(3.1)
5.5(0.2)
11(1)
19.8(0.8)
37.1(0.2)
5b
1,4-(MeN)218C6
KCl
4.71(0.54)
8.44(0.54)
1:2
[e]
5c
6
5,18-Dioxo-1,4-N218C6
15C5
KCl
KCl
1:1
1:2
1:1
1:2
1:1
1:2
1:1
1:2
1:1
3.25(0.03)
5.75(0.02)
3.46(0.05)
5.95(0.05)
3.40(0.03)[d]
5.83(0.05)[d]
3.35[9e]
18.6(0.2)
32.8(0.1)
19.8(0.3)
34.0(0.3)
19.4(0.2)[d]
33.3(0.3)[d]
19.1[9e]
30.1(0.2)
68.6(0.2)
29.6(0.3)
66.5(0.5)
Ϫ11.5 (0.3)
Ϫ35.8(0.2)
Ϫ9.8(0.4)
KI
Ϫ32.5(0.6)
KI
KI
32.6[9e]
Ϫ13.5[9e]
Ϫ35.2[9e]
0.0(0.5)
9.4(1.3)
6.00[9e]
34.2[9e]
69.4[9e]
7a
1,7-(MeN)215C5
KCl
3.66(0.07)
5.98(0.20)
20.9(0.4)
34.1(1.1)
20.9(0.3)
24.7(0.7)
1:2
[e]
7b
8a
8b
8c
2,6-Dioxo-1,7-N215C5
1,4-N215C5
KCl
KCl
KCl
KCl
[e]
1,4-(MeN)215C5
1:1
1:1
1:2
1:1
1:1
1:1
1:1
1.93(0.02)
1.5
4.8(0.8)
10.49[9b]
10.41[9f]
4.22[9g]
1.60(0.02)
11.0(0.1)
8.6
27.4(4.5)
59.9[9b]
59.4[9f]
24.1[9g]
9.1(0.1)
15.1(0.4)
-10
Ϫ4.1(0.4)
19
5,15-Dioxo-1,4-N215C5
-6.5(1.2)
33.8(4.6)
222
[222]
Kϩ
75.0[9b]
Ϫ15.1[9b]
Kϩ
9
10
21C7
1,10-N221C7
Kϩ
35.9[9g]
13.8(0.5)
Ϫ11.8[9g]
Ϫ4.7(0.5)
KNO3
[a]
[b]
For experimental conditions and details see text. Ϫ Complexation data are given for equilibria: 1:1: Mϩ ϩ L ϭ [ML]ϩ, 1:2: Mϩ
ϩ
2L ϭ [ML2]ϩ , where applicable, see text. Ϫ Uncertainties are given as standard deviations in brackets. Ϫ Potentiometric titration
with solid state potassium selective electrode. ionic strength: 0.05  Et4NI. Ϫ Too low heat effect for calorimetry.
[c]
[d]
[e]
Results and Discussion
ameter TIP3 water box with 428 water molecules yielded
two distinct energy minima for the ligand 2a, the one with
The free energy ∆G of complexation of the ligand two axially oriented lone pairs (conformer 2aa) at the nitro-
1,10DA18C6 (2a) with potassium (Table 1) as with sodium gen atoms being 9 kJ/mol more stable than the one with
(Table 2) is only about 1/3 of that of 18C6 (1), which com- two equatorial lone pairs (2ee) (Scheme 2). Dissection of
pared to the values known for the cryptand [2.2.2] is unex- energy terms from the CHARMm minimisation (Table 3)
pected on the basis of number and electron donating ca- as well as AM1 calculations[11] show that the destabilisation
pacity of the heteroatoms. Computer aided molecular mod- of 2ee compared to 2aa is almost entirely due to electro-
1380
Eur. J. Org. Chem. 1998, 1379Ϫ1389

Aza Crown Ether Complexes
Scheme 1. Structures of ligands, with Ϫ∆G values Ϫin italicsϪ [kJ/mol] of complexation with Kϩ in methanol (other values see Table 1).
FULL PAPER
Eur. J. Org. Chem. 1998, 1379Ϫ1389
1381

H.-J. Schneider
FULL PAPER
static repulsion between the ee lone pairs at the nitrogen towards metal cations one should be able to revert this by
and the electron pairs of the ring oxygen atoms. In contrast, exchanging the NϪH hydrogen atoms in 2a by methyl
the force field calculations show in agreement with litera- groups. The large methyl substituents force the lone pairs
ture computations[12] that the cryptand 222 predominates into the ee conformation necessary for optimal complex sta-
(vide infra) in the in-in lone pair conformation, thus en- bility. It is gratifying, that the experiments (Table 1) support
abling a maximum interaction with the metal cation.
this mechanistic reasoning. In line with this, methylation
Table 2. Stability constants (lgβ units) and thermodynamic parameters (kJ/mol) for complexation of azacrown ethers with sodium cation
in methanol at 298 K.[a]
No
Ligand
Salt
M:L[b]
lgβ[c]
Ϫ∆G
Ϫ∆H
T∆S
1
18C6
NaI
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:1
1:2
1:1
1:1
1:1
1:2
1:1
4.36(0.02)[d]
1.63(0.02)[d]
3.53(0.08)[d]
1.54(0.05)
2.78(0.06)
1.96(0.12)[d]
4.01(0.08)
3.78(0.06)[d]
3.93[9j]
24.9(0.1)[d]
9.3(0.1)[d]
20.1(0.5)[d]
8.8(0.3)
15.9(0.3)
11.2(0.7)[d]
22.9(0.5)
21.6(0.3)[d]
22.4[9j]
2a
2b
2c
3a
1,10-N218C6
1,10-(MeN)218C6
1-BzN-10-N18C6
N18C6
NaI
NaI
NaCl
NaNO3
NaNO3
NaOH
NaI
15.8(1.2)
10.5(0.3)
Ϫ7.0(1.2)
5.4(0.4)
3b
MeN18C6
19.5(0.4)
3.4(0.6)
Naϩ
3c
BzN18C6
NaNO3
3.02(0.06)
6.23(0.04)
2.4(0.4)
17.2(0.3)
35.6(0.2)
13.7(2.3)
4
27.9(0.8)
47.4(0.5)
Ϫ1.2(0.2)
ENDO
Ϫ15
Ϫ10.7(0.9)
Ϫ11.8(0.5)
15
4a
4c
5c
1,7-N218C6
8,18-Dioxo-1,7-N218C6
5,18-Dioxo-1,4-N218C6
NaCl
NaCl
NaCl
0.7
1.1
6.3
21
34
2.8
16
Ϫ18
6
15C5
NaI
3.25(0.02)[d]
5.18(0.04)[d]
18.5(0.1)[d]
29.6(0.2)[d]
1:2
[e]
7b
2,6-Dioxo-1,7-N2Ϫ15C5
NaCl
[a]
[b]
For experimental conditions and details see text. Ϫ Complexation data are given for equilibria: 1:1: Mϩ ϩ L ϭ [ML]ϩ, 1:2: Mϩ
ϩ
2 L ϭ [ML2]ϩ , where applicable, see text. Ϫ Uncertainties are given as standard deviations in brackets. Ϫ Potentiometric titration
with solid state potassium selective electrode; ionic strength: 0.05  Et4NI. Ϫ Heat effect too low for calorimetry.
[c]
[d]
[e]
Scheme 2a. Force field calculated conformation of 1,10DA-18C6
2a: A with axial position of lone pairs; B with equato-
rial positions of lone pairs.
Table 3. Comparison of individual energy contributions from
CHARMm-force field calculations of 1,10DA18C6 2a.[a]
1,10DA18C6-
LPax,ax
1,10DA18C6-
LPeq,eq
Total CHARMm energy
Bond energy
Ϫ9.834
0.360
Ϫ0.927
2.995
Angle energy
2.900
0.371
Dihedral energy
Inproper energy
2.385
0.000
2.838
0.000
Lennard-Jones energy
Electrostatic energy
Constraints, other
Ϫ31.115
15.663
0.000
Ϫ30.368
23.236
0.000
[a]
All data in [kJ/mol].
Scheme 2b. LPin,in Conformation of Cryptand [222]
leads to an substantial ∆G increase also for the mono aza
crown ether complex 3a. Similar increases, likely all due to
stabilisation of in, or equatorial, orientation of the N lone
pairs by alkyl groups at the nitrogen atoms are seen with
other ligands such as 2c. It should be noted that the in-
creased stability of the N-alkyl crown ether complexes can-
not be due the a higher electron donor capacity of these
groups, which does not change within standard deviation
limits of Ed and Cd upon alkylation.[4] The higher basicity
of nitrogen compared to oxygen should lead to stronger hy-
drogen bonds with protic solvents; therefore one should ex-
pect stronger complexation increase in non-protic solvents
than observed with normal all-oxygen crown ethers. In ab-
sence of reliable literature data on aza crown ether com-
If the unfavourable aa lone pair orientation in the diaza plexes in solvents like acetonitrile, chloroform etc this needs
crown ether 2a is essentially responsible for its low affinity further investigation.
1382
Eur. J. Org. Chem. 1998, 1379Ϫ1389

Aza Crown Ether Complexes
FULL PAPER
The experimentally observed binding energies ∆G for the in methanol prior to complexation. NMR titrations of
MeϪN crown ether complexes (2b, 3b) are, however, still 1,10DA18C62a with different potassium salts (KCl, KOH)
below the ones predicted from additive increments, in sharp were carried out in [D4]MeOH with TMS as internal stand-
contrast to the cryptand 222 complexes (Table 4) The rea- ard at 298 K. Practically there was no difference observed
son for this becomes evident from the CHARMm Ϫ calcu- in the titration curves with the two different potassium salts.
lated distances between the donor heteroatoms to the cen- The NMR shifts in acidic medium (50 eq. of DCl in
tral potassium ion. These are after energy minimisation of [D4]MeOH) for the two observed CH2 groups were: δ (α-
˚
the corresponding complexes on the average 0.2 A larger in CH2) ϭ 3.29 and δ (β-CH2) ϭ 3.77 ppm. The shifts in pure
the diaza crown ether complexes compared to those in the [D4]MeOH were: δ (α -CH2) ϭ 2.74 and δ (β-CH2) ϭ 3.56.
cryptate. At the same time the ion cannot fully immerse Therefore, the observed chemical shifts during the titration
into the 2ee cavity, again in contrast to the cryptate (Fig- with potassium salts of the α- and β-methylene groups (ad-
ure 1).
jacent to nitrogen) can only be due to complexation by the
metal ion. This titration then is one of the few examples
were one can follow a crown ether complexation by NMR,
since normal all-oxygen crown ethers show no significant
shift changes upon complexation. Association constants in
agreement with the calorimetric data were obtained by non-
linear least square fitting to a 1:1 complexation model (see
Table 1 and Figure 2).
Table 4. Predicted and experimental (calorimetry) complexation
free energies ∆G of Kϩ-azacrown and Kϩ-cryptand [222] comple-
xes.[a]
[b]
[c]
Crown
∆Gexp
∆Ginc
∆∆G
N18C6 3a
Ϫ10.6
Ϫ31.1
Ϫ10.0
Ϫ29.5
Ϫ60.0
Ϫ35.5
Ϫ40.4[d]
Ϫ40.4
29.7
9.2
36.2
16.7
Ϫ2.7
Ϫ1.0
MeN18C6 3b
1,10-N218C6 2a
1,10-(MeN)218C6 2b
[222] 222
Ϫ46.2[d]
Ϫ46.2
The observed thermodynamic parameters are for most
cases in line with the essentially enthalpic arguments dis-
cussed above. Thus, the low stability of aza crown ether
complexes with NH compared to he all-oxygen analogs is
indeed largely due to ∆H disadvantages (cf. for instance 2a
or 3a vs. 1; or 10 vs. 9). Also, the stabilising effect of substi-
tution in the NR ligand complexes (2b, 2c, 4b) largely shows
up in ∆H advantages. Striking differences are seen es-
pecially with the 1,7-diaza-18-crown-6 4a, where com-
plexation is entirely driven by T∆S. Similar, although less
pronounced favourable T∆S contributions are also visible
for the other aza crown ethers, even more for the observed
[ML2]ϩ complexes. This is in sharp contrast to ligands such
as 18C6 or 15C5 itself, where complexation is accompanied
by T∆S disadvantage which increases with the affinity, ob-
viously as consequence of the then more limited mobility.
In line with the arguments above, a similar situation with
Ϫ although less pronounced Ϫ entropic disadvantage is ob-
served with the N-substituted aza crown ethers such as 2b
or 3c, which are again better disposed for stronger com-
plexation.
Ϫ57.3
Ϫ34.5
18C6 1
[a] In [kJ/mol]. Ϫ [b] Measured in MeOH. Ϫ [c] Calculated incremen-
tally, see text. Ϫ
reference.
Data for the Kϩ-18C6 complex are given as
[d]
Figure 1. Geometries of the 1,10DA 18C6-LPax,ax- (2a), 1,10DA
18C6-LPeq,eq- (2a) and cryptand [2.2.2]-potassium complexes, from
CHARMm calculations.
Inoue et al.[13] have assembled many data on entropyϪ
enthalpy compensations with ionophor complexes. Their
correlations (for a selection see Table 5) suggest, that less
preorganised ligands like glymes suffer from large entropy
disadvantage which needs to be completely compensated by
enthalpy contributions. This is less so for cryptand and
crown ether complexes, which suffer only by 44% or 77%
entropic disadvantage. Although the underlying isoequilib-
rium correlations seem to be of sufficient statistical quality
one should keep in mind the danger of systematic errors in
such correlations, particularly if they are based on tempera-
ture dependent equilibrium and not on calorimetric meas-
urements. Inoue et al already noted some incongruencies
with natural ionophores, which unexpectedly showed slopes
of α ϭ 0.93, closer to glymes or podands than to crown
ethers.
Calorimetric measurements were carried out with some
other aza crown ethers (Scheme 1, Table 1). They show
again relatively large ∆G values in comparison to the all-
oxygen analogs as soon as the nitrogen atoms bear alkyl
groups (cf. e.g., 4a/b, 5a/b; 7a/b, 8a/b). Vicinal nitrogen
atoms as in the 1,4-diaza-18-crown-6 5a present a special
case in view of the high pKA for the first, and the low pKA
for the second step of the protonation.
NMR shift data as well as titrations (Figure 2) show,
Typically, most crown ethers also are expected to loose
however, that all aza crown ether ligands are not protonated little conformational freedom upon complexation; the still
Eur. J. Org. Chem. 1998, 1379Ϫ1389
1383

H.-J. Schneider
FULL PAPER
Figure 2. Experimental NMR shifts (dots) and fitted line for the a larger role in crown ethers as compared to cryptands. Our
NMR titration of 1,10-diaza 18C6 (2a) with KOH in [D4]MeOH
own results with aza crown ether complexes (Table 5),
at 298 K. Association constant: 93ϩ/Ϫ8; ∆G ϭ Ϫ11.2 kJ molϪ1
;
which do not suffer from the error problems involved with
the temperature dependent equilibrium measurements
based on the vanЈt Hoff method, and are based on data in
only one solvent, also show a satisfactory linear isoequilib-
rium correlation (Figure 3), although there are outliers
which cannot be rationalised at this time. The observed
slopes and intercepts (Table 5) are between those observed
of crown and cryptand complexes, indicating entropic dis-
advantages which are compensated to about 65% by en-
thalpic contributions.
shifts at 100% complexation from fit: CIS (α) ϭ 0.024 ppm, CIS
(β) ϭ 0.064 ppm.
Figure 3. Correlation between ∆H and T∆S for the complexation
of Kϩ with aza crown ethers, aza oxo crown ethers (amide crowns),
18-crown-6 and 15-crown-5 in MeOH for the Kϩ ϩ L v (Kϩ)L
equilibrium: T∆S 13.2(3.9) ϩ 0.65(0.15) ∆H, N ϭ 17, R ϭ 0.923,
F ϭ 86.5, SD ϭ 4.8.
Kϩ-[222]
Kϩ-1,10DA18C6LPeq,eq
The cyclic amides 4c, 5c, 7b and 8c were available as syn-
thetic precursors and were subjected to stability measure-
ments. In natural ionophores such as valinomycin the nega-
tive partial charges of the oxygen atom in the carbonyl
group of amide or ester functions can also contribute to
binding, requiring, however, orientation of the CϭO units
towards the metal. In small cyclic amides this would induce
large strain in the ligands in view of the strongly preferred
transoid R-CO-NH-RЈ conformation.[14] Moreover amide
nitrogen have essentially lower electron donor capacity
(Ed ϭ 0.23 and Cd ϭ 0.32), which is 8Ϫ10 times lower than
those in aza crown ethers (Ed ϭ 2.78 and Cd ϭ 2.60) and
in alkylϪaza crown ethers (Ed ϭ 2.34 and Cd ϭ 2.57).[4]
Thus, the amide crown ethers have in fact only four coordi-
nation centers as 12-crown-4, which are available for the
complexation with a metal cation.
Table 5. Entropy-enthalpy compensations: T∆S vs ∆H [kJ/mol
units]; (from Table 1, Figure 4, and ref.[17]).
Complexes
N[a] r[b]Ͻ/
T∆S[d]
en-
try'<εντρψ>α[c]
all aza crown ethers
complexes
27
0.828
0.59±0.16
17±6
only 1:1 cplxs
17
9
0.923
0.958
0.88
0.98
0.98
0.65
0.65±0.15
0.81±0.22
0.77±0.02
0.94
1.02±0.02
0.44±0.04
13±4
32±8
11.7±0.4
13.4±0.8
15.5±0.8
14.6±1.7
only 1:2 cplxs
crowns 1:1 cplxs[15]
crowns 1:2 cplxs[15]
glymes/podands[15]
cryptands[15]
598
85
151
160
[a]
[b]
Number of points used. Ϫ
Linear correlation coefficient. Ϫ
[c]
[d]
Slope. Ϫ Intercept.
high slopes observed by Inoue et al.[13] (α ϭ 0.88) might
The complexation energies are indeed smaller than the
well be due to solvation effects which are expected to play one of the analog without oxo group if we compare them
1384
Eur. J. Org. Chem. 1998, 1379Ϫ1389

Aza Crown Ether Complexes
FULL PAPER
Figure 4. Quanta/CHARMm-minimized structures of amide-crown to considerable built up of strain energy in the ligand. In
structures 5c and 8c and their Kϩ complexes.
consequence the cation can only be in poor contact, and
only with 5 out of the 6 oxygen atoms, as visible in the
corresponding CPK model (Figure 4), and the com-
plexation is below the calorimetric detection limit. With the
smaller macrocycle 8c both carbonyl oxygens can be in con-
tact with the cation Ϫ again after inducing considerable
conformational change; this according to the molecular
mechanics calculations costs, however, much less energy
than in the case of 5c. Even though in the optimal con-
former of the 8c Kϩ complex (Figure 4) the ether oxygen
˚
atoms are far too distant (with d around 6 A) and lead
to an outside location of the cation (see Figure 4), both
electronegative carbonyl oxygen atoms can contact the cat-
ion and thus allow a moderately stable complex, based
mostly on an enthalpic contribution with small entropic
disadvantage (Table 1).
Measurements of the complexes of aza crown ethers with
sodium salts (Table 2) showed similar to the potassium
complexes the strong ∆G increase with alkylation of the ni-
trogen atoms, indicating the same stereoelectronic effect as
discussed above. The different range of stabilities observed
with the amide-type ligands 4c and 5c are expected as
consequence of the different ion diameters of Na compared
to K, requiring a smaller degree of conformation and thus
strain changes for complexation.
to the aza crown ether complexes without alkyl groups at
nitrogen. The disadvantage of the NH containing ligands
due to the predominant axial lone pair orientation does of
Conclusions
The data show how simple predictions of ionophore sta-
bilities on the basis of additive increments for the interac-
tion between ligand element and metal cation are limited
by non-ideal conformations in the ligands. Computer aided
molecular modelling provides a convenient tool to control
the necessary conformational requirements and can be of
great help not only in the understanding of measured bind-
ing energy differences, but also in the de novo design of syn-
thetic ionophores. The results stress the significance of lone
pair orientation in ionophores, which until now is not ex-
plicitly included in most of the available force fields.
course not exist in the amides. Ligand 8c forms a remark-
ably strong 1:2 [ML2]ϩ complex as well as a moderately
strong 1:1 complex. Preliminary molecular mechanics cal-
culations show for the energy minimised Kϩ 5c complex
with all-transoid amid conformations simultaneous interac-
tions of the cation with 5 of 6 oxygen atoms, except with
one of the amid oxygens (Figure 4), however with distances
K...O much smaller than those in the ideal crown complex
Kϩ18C6 (see Table 6). In addition, severe torsional changes
at least around one angle are necessary to bring at least one
of the two carbonyl oxygens close to the cation, leading
We thank the Volkswagen-Stiftung, Hannover for a generous
grant for these studies. The work in Saarbrücken is partially also
supported by the Deutsche Forschungsgemeinschaft, Bonn, and the
Fonds der Chemischen Industrie, Frankfurt. The work in Chernogo-
lovka is also supported by the INTAS grant 94Ϫ1914.
Table 6. Dihedral angles and Kϩ...O distances in the amide crowns
5c and 8c and their Kϩ complexes.[a]
˚
Dihedral Angles
5c
5c/Kϩ
Distances [A]
5c/Kϩ
4Ϫ5Ϫ6Ϫ7
53.4
157.3
Ϫ55.8
53.7
Ϫ57.6
163.7
Ϫ75.3
5-O
7
4.4
3.2
3.0
3.2
3.1
2.9
Experimental Section
7Ϫ8Ϫ9Ϫ10
Ϫ63.8
66.2
10Ϫ11Ϫ12Ϫ13
13Ϫ14Ϫ15Ϫ16
16Ϫ17Ϫ18Ϫ1
1Ϫ2Ϫ3Ϫ4
10
Calorimetric Measurements: Determination of logβ and ∆H
complexation values were carried out using the calorimetric ti-
tration technique at 298.15 K as described before.[5][6] The reaction
heats were measured with a microcalorimetry system LKB (Model
2107/112). Total (overall) concentrations of reagents which were
used in calorimetric and potentiometric measurements are given in
Table 7.
Ϫ64.0
52.9
Ϫ63.9
13
16
18-O
˚
Dihedral Angles
8c
8c/Kϩ
Distances [A]
8c/Kϩ
4Ϫ5Ϫ6Ϫ7
Ϫ20.3
Ϫ51.0
53.6
Ϫ43.0
Ϫ51.6
41.6
50.5
Ϫ52.0
Ϫ35.7
Ϫ48.9
5-O
7
3.0
5.6
6.6
5.6
3.0
7Ϫ8Ϫ9Ϫ10
10Ϫ11Ϫ12Ϫ13
13Ϫ14Ϫ15Ϫ1
1Ϫ2Ϫ3Ϫ4
The reliability of the calorimetric determinations was checked by
measuring stability constants and enthalpy changes of the 18-
crown-6 complexation with barium cation in water at 298K, yield-
ing excellent agreement of our data with known logK values, and
satisfactory agreement for enthalpies (Table 8).
10
13
15-O
[a]
From CHARMm force field calculations.
Eur. J. Org. Chem. 1998, 1379Ϫ1389
1385

H.-J. Schneider
FULL PAPER
Non-linear least square fitting for the calculation of stability con-
stants and enthalpy changes from experimental calorimetric and
potentiometric titrations was performed with a new and extended
32-bit PC version of the general multifit program CHEM EQUI
for WINDOWS 95/NT, which itself was described earlier[15a][15b]
(available from Dr. V. SolovЈev).
Table 7. Total (overall) concentrations of reagents used in calorime-
tric and potentiometric measurements; CoS, total concentration of
a salt; Co , total concentration of ligand.
L
Ligand
18C6
Salt
Co
Co
S
L
Mmol/l
mmol/l
The stoichiometries (compositions) of the complexes were calcu-
lated for series of nMϩ ϩ ml ϭ [MnLm]nϩ equilibrium models, tak-
ing into account variable number and combinations of complexes
in a solution (Mϩ ϭ metal cation, L ϭ ligand; n, m ϭ 1, 2 and 3).
The models given in Tables 1 and 2 are those for which a satisfac-
tory fit of experimental data of several calorimetric titrations was
obtained on the basis of statistical criteria. These were mainly the
Hamilton R factor[16a][16b] and residuals (Qexp. Ϫ Qcalcd.) analysis
for fitness test,[16c][16d] were Qexp. and Qcalcd. are calorimetric ti-
tration heats. Systematic deviations in the fitting to 1:1 models
(complex [ML]ϩ) were frequently observed, in which cases satisfac-
tory fit with Hamilton R-factors below 2% were obtained by evalu-
ation of additional [ML2]ϩ complexes (see Tables 1 and 2). Avail-
able literature lgβ and ∆H values for the complexation of aza-18-
crown-6, 1,10-diaza-18-crown-6, and N,N dimethyl-1,10-diaza-18-
crown-6 with potassium cation in MeOH agreed with values ob-
tained in this work (Tables 1 and 2).
1
KI [a]
KNCS
7.9Ϫ9.3
11.2Ϫ11.8 1.1Ϫ11.1
1.4Ϫ16.1
0.8Ϫ14.9
7.3Ϫ8.3
1.5Ϫ17.9
7.3Ϫ7.9
7.3Ϫ7.8
0.4Ϫ5.2
0.7Ϫ5.2
6.9Ϫ7.9
7.0Ϫ7.9
0.8Ϫ10.2
0.8Ϫ10.2
0.4Ϫ4.7
0.5Ϫ3.0
6.4Ϫ6.9
0.5Ϫ6.1
0.5Ϫ4.7
11.7Ϫ12.9
1.3Ϫ15.7
12.0Ϫ12.8
1.3Ϫ10.5
2.5Ϫ10.4
2.7Ϫ4.9
4.2Ϫ4.5
1.4Ϫ17.6
1.3Ϫ15.5
4.3Ϫ4.6
5.4Ϫ6.2
2.7Ϫ2.9
4.2Ϫ4.5
1.0Ϫ8.7
4.2Ϫ4.5
4.0Ϫ4.2
0.8Ϫ13.5
1.2Ϫ13.6
1.3Ϫ15.2
4.6Ϫ5.1
1.3Ϫ15.5
4.0Ϫ4.5
0.8Ϫ13.5
4.3Ϫ4.6
5.4Ϫ6.1
4.3Ϫ4.6
4.0Ϫ4.5
5.4Ϫ6.1
4.0Ϫ4.6
4.3Ϫ4.6
4.0Ϫ4.1
5.4Ϫ6.1
6.6Ϫ7.1
5.8Ϫ6.2
3.3Ϫ27.1
1.3Ϫ16.6
5.1Ϫ5.8
5.0Ϫ5.3
7.5Ϫ8.0
5.0Ϫ5.5
5.1Ϫ5.9
4.0Ϫ4.5
6.0Ϫ6.8
NaI[a]
KOH
KI[a]
2a 1,10-N218C6
NaI[a]
KCl
2b 1,10-(MeN)218C6
KOH
KI[a]
NaI[a]
KCl
2c 1-(BzN)-10-N18C6
NaCl
KCl
2d 1,10-(BzN)218C6
2e 1,10-(AcN)218C6
KOH
KI[a]
3a N18C6
KNO3
KOH
NaNO3[a] 6.7Ϫ7.3
NaNO3 8.9Ϫ9.8
3b MeN18C6
KI[a]
NaOH
NaI[a]
KNO3
7.0Ϫ8.0
0.8Ϫ8.6
6.9Ϫ7.8
0.5Ϫ11.5
Typical calorimetric titration curves with fitting are shown in
Figures 5aϪc. For controlling the possible influence of aza crown
ether protonation we measured complexation of 2b and 3a not only
with potassium salts KA (A ϭ IϪ, ClϪ, NO3Ϫ), but also with pot-
assium hydroxide KOH. In line with NMR measurements (see be-
low) the negligible difference between the KA and the KOH experi-
ment shows, that at the low proton concentration in pure methanol
the protonation of aza crown ethers is negligible.
3c BzN18C6
NaNO3 7.0Ϫ7.7
4a 1,7-N218C6
KCl
0.8Ϫ6.0
NaCl
KCl
1.6Ϫ12.6
0.5Ϫ6.2
4b 1,7-(MeN)218C6
4c 8,18-Dioxo-1,7-N218C6
KCl
1.1Ϫ8.5
NaCl
KCl
1.1Ϫ8.5
5a 1,4-N218C6
5b 1,4-(MeN)218C6
5c 5,18-Dioxo-1,4-N218C6
0.8Ϫ10.2
0.8Ϫ10.2
1.5Ϫ10.0
1.5Ϫ12.2
1.6Ϫ21.1
1.6Ϫ21.1
7.8Ϫ9.3
KCl
Potentiometric measurements were carried out with using the ti-
tration technique at 298.15 K as described before,[5][6] using solid
state potassium and sodium selective electrodes and a reference Ag/
Agϩ electrode. Ionic strength of the solutions was kept constant at
0.05  Et4NI in all experiments.
KCl
NaCl
KCl
6
15C5
KI
KI[a]
NaI[a]
KCl
6.4Ϫ7.4
7a 1,7-(MeN)215C5
7b 2,6-Dioxo-1,7-N215C5
0.8Ϫ10.4
0.4Ϫ10.0
0.4Ϫ10.0
0.5Ϫ10.0
0.5Ϫ12.9
1.5Ϫ12.5
1.0Ϫ11.5
NMR titrations were carried out on a Bruker AM400 or
DRX500 system at 298 K. The concentrations and ratios of ob-
served and added compounds in a titration were chosen to pass a
range of complexation between 20 and 80%. Evaluation of the ob-
tained data was done by non-linear least square fitting with the
program SigmaPlot (Jandel Scientific) using equations for 1:1 host/
guest complexation.[17]
Reagents: The salts KX and NaX (X ϭ ClϪ, IϪ, SCNϪ, OHϪ,
NO3Ϫ) were either commercially available and used without further
purification, or were purified as described.[5][6] Methanol was recti-
fied over magnesium methanolate under nitrogen.
KCl
NaCl
KCl
8a 1,4-N215C5
8b 1,4-(MeN)215C5
8c 5,15-Dioxo-1,4-N215C5
10 1,10-N221C7
KCl
KCl
KNO3
[a]
Potentiometric titration.
With the aza crown ethers average standard deviations were less
than 0.25 in lgK, 0.6 kJ/mol in ∆H for the Mϩ ϩ L ϭ [ML]ϩ (M ϭ
K, Na) equilibria. Standard deviations (in brackets) of lgK and ∆H
values for the aza crown ethers were a little higher in comparison
to normal crown ethers.[5][6] This can be due to often lower lgK
Ligands: The structures of all new compounds were secured by
elemental analysis and 1H spectra. NMR spectra were recorded on
Bruker CXP200 spectrometer in CDCl3 with tetramethylsilane as
values, to ∆H values near zero, or also to slower complexation with internal reference. Melting points (uncorrected) were measured on
aza crown ethers. a Boetius PHMK-05 instrument.
Table 8. Stability constant (lgK units) and thermodynamic parameters (kJ/mol) for complexation of 18-crown-6 with barium cation in
water at 298 K.
Ligand
Salt
Solvent
Reaction
lgK
∆H, kJ/mol
Ref.
18-crown-6
18-crown-6
BaCl2
BaCl2
H2O
H2O
MϩϩL ϭ [ML]ϩ
MϩϩL ϭ [ML]ϩ
3.80 (0.03)
3.87
Ϫ29.7(0.5)
Ϫ31.7
this work
[19a,19b]
1386
Eur. J. Org. Chem. 1998, 1379Ϫ1389

Aza Crown Ether Complexes
Figure 5a. Calorimetric titration curve for the complexation of Figure 5c. Calorimetric titration curve for the complexation of
FULL PAPER
1,10-diaza-18C6 2a with KOH in MeOH at 298 K. Experimental
interaction heats (cycles) and calculated titration curve vs. the re-
agents concentration ratio CoL/CoM. Experimental data: the ligand
1,10-diaza-21-crown-7 10 with KNO3 in MeOH at 298 K. Experi-
mental interaction heats (cycles) and calculated titration curve vs.
the reagents concentration ratio CoL/CoM. Experimental data: li-
solution (CoL 80.7 m and ∆V 0.0212 ml for each point) was added gand solution (Co 95.1 m and ∆V 0.0212 ml for each point) was
into the mixture (Co 5.8 m, Co 0.84 m, Vo 2.02 ml).
added to the titraLted mixture (Co 6.8 m, Co 1.0 m, Vo 2.02
M
L
M
L
ml).
of formic acid. The mixture was heated to 80Ϫ90°C or 24 h. After
cooling to room temperature solid KOH was added until pH ϭ 11
was reached; the mixture was extracted by CHCl3 (3 ϫ 5 ml) and
the solvent was removed in vacuo. The residual oil was purified
twice by distillation in vacuo. Yield 2.26 g (64%); b.p. 140Ϫ142°C/
0.9 Torr. Ϫ C14H30N2O4; mol. mass 290.41; calcd. C 57.9, H 10.4,
Figure 5b. Calorimetric titration curve for the complexation of 1,7-
di(methylaza)-15-crown-5 4b with KCl in MeOH at 298 K. Experi-
mental interaction heats (cycles) and calculated titration curve vs.
the reagents concentration ratio CoL/CoM. Experimental data: the
ligand solution (Co 277 m and ∆V 0.0106 ml for each point)
added to the mixtLure (Co 12.8 m, Co 1.5 m, Vo 2.01 ml).
1
N 9.6; found C 57.8, 57.7; H 10.2, 10.2; N 9.5, 9.6. Ϫ H NMR:
δ ϭ 2.31 (s, 6 H, 2 NCH3 ), 2.70 (m, 8 H, 4 OCH2N), 3.62 (m, 12
M
L
H, 2 OCH2).
N-Benzyl-1,4,10,13,-tetraoxa-7,16-diazacyclooctadecane (2c): 5.0
g (11.2 mmol) commercially available N,N-dibenzyl-1,4,10,13,-
tetraoxa-7,16-diazacyclooctadecane were dissolved in 450 ml of dry
ethanol. Palladium on charcoal (1.0 g) and 8 ml of glacial acetic
acid were added and the mixture was hydrogenated for about 30 h
at room temperature under atmospheric pressure. The composition
of the reaction mixture was controlled by TLC and 1H NMR. The
reaction mixture was filtered and the solvent was evaporated in
vacuo. 100 ml 15% NaOH were added and the mixture was ex-
tracted with CHCl3 (7 ϫ 40 ml). The extract was washed with water
(2 ϫ 30 ml) and evaporated in vacuo. The crude residue was puri-
fied by column chromatography [Al2O3/CHCl3, CHCl3/EtOH
(20:1)]. Ϫ Yield 1.7 g (43%) as oil. Ϫ C19H32N2O4; mol. mass
352.47; calcd. C 64.8, H 9.2, N 8.0; found: C 64.6, 64.5; H 9.4, 9.3;
N 7.9, 8.0. Ϫ 1H NMR (CCl4): δ ϭ 2.20 (br. s, 1 H, NH), 2.70 (m,
8 H, 4 CH2N), 3.48 (m, 16 H, 2 OCH2), 3.64 (s, 2 H, NCH2Ph),
7.26 (m, 5 H, ArϪH).
1,4,7,10,13-Pentaoxa-16-azacyclooctadecane (3a) was prepared as
described previously[18b]
.
N-Methyl-1,4,7,10,13,-pentaoxa-16-azacyclodecane (3b) was ob-
tained analogous to 2b from 2.0 g (7.6 mmol) of 3a, 0.54 g (18.2
mmol) formaldehyde (as 30% aqueous solution) and 3.5 g (76.0
mmol) of formic acid, 80Ϫ90°C, 24 h. Ϫ Yield 0.7 g (53%); b.p.
120Ϫ122°C/0.9 Torr, ref.[18b] 160°C/0.03 Torr. Ϫ C13H27NO5; mol.
Synthesis of Ligands: 1,4,7,10,13,16-Hexaoxacyclooctadecane (1)
was commercially available, and purified[5] before use.
1,4,13,18-Tetraoxa-7,16-diazacyclooctadecane (2a) was synthe-
sized as described previously[18a]
.
N,N-Dimethyl-1,4,10,13,-tetraoxa-7,16-diazacyclooctadecane (2b): mass 277.36; calcd. C 56.3, H 9.8, N 5.0; found C 56.5, 56.4; H
1
To a solution of 3.2 g (12.2 mmol) 2a in 0.89 g (29.7 mmol) formal-
10.0, 10.1; N 5.0, 5.0. Ϫ H NMR: δ ϭ 2.30 (s, 3 H, NCH3), 2.69
dehyde (as 30% aqueous solution) were added 5.6 g (122.0 mmol) (m, 4 H, 2 CH2N), 3.64 (m, 20 H, 10 OCH2).
Eur. J. Org. Chem. 1998, 1379Ϫ1389
1387

H.-J. Schneider
FULL PAPER
N-Benzyl-1,4,7,10,13,-pentaoxa-16-azacyclodecane (3c) was ob-
58.5, H 10.6, N 11.4; found C 58.3, 58.4; H 10.7, 10.8; N 11.3,
11.4. Ϫ 1H NMR: δ ϭ 2.32 (s, 6 H, 2 NCH3 ), 2.72 (m, 8 H, 4
OCH2N), 3.64 (m, 12 H, 2 OCH2).
tained as described before.[18f]
1,13-Diaza-4,7,10,16-tetraoxacyclooctadecane (4a) was obtained
by reduction of 1.6 g (6.5 mmol) 4c with 1.6 g (40 mmol) lithium
aluminium hydride in tetrahydrofuran under reflux during 24 h. Ϫ
Yield 2.3 g (49%). Ϫ Colourless crystals, m.p. 89Ϫ90°C (petroleum
ether), ref.[18g] m.p. 89Ϫ90°C.
1,7-Diaza-4,10,13-trioxacyclopentadeca-2,6-dione (7b) was ob-
tained as described previously.[18e]
1,13-Diaza-4,7,10,-trioxacyclopentadecane (8a) was prepared
analogous to 4a by reduction of 1.1 g (4.5 mmol) of 8c with 2.0 g
(52.6 mmol) of lithium aluminium hydride in boiling tetrahydrofu-
ran during 24 h. Ϫ Yield 0.58 g (60%); b.p. 123Ϫ125°C/0.9 Torr.
Ϫ C10H22N2O3; calcd. C 55.0, H 10.2, N 12.8; found C 54.8, 54.9;
N,N-Dimethyl-4,7,10,16-tetraoxa-1,13,-diazacyclooctadecane (4b)
was obtained analogous to 2b from 2.0 g (7.6 mmol) of 4a, 0.54 g
(18.2 mmol) formaldehyde (as 30% aqueous solution) and 3.5 g
(76.0 mmol) of formic acid, 80Ϫ90°C, 24 h. Ϫ Yield 0.7 g (53%);
b.p. 120Ϫ122°C/0.9 Torr, ref.[18c] 160°C/0.03 Torr. Ϫ C14H30N2O4;
mol. mass 290.41; calcd. C 57.9, H 10.4, N 9.6; found C 57.8, 57.7;
1
3
H 10.0, 10.1; N 12.6, 12.7. Ϫ H NMR: δ ϭ 1.00 (t, J ϭ 7.0 2 H,
2 NH ), 2.62 (m, 4 H, NCH2CH2N), 2.80 (m, 4 H, OCH2CH2N),
3.62 (m, 12 H, 2 OCH2).
1
H 10.3, 10.3; N 9.6, 9.7. Ϫ H NMR: δ ϭ 2.32 (s, 6 H, 2 NCH3),
N,N-Dimethyl-4,7,10-triaoxa-1,13-diazacyclooctadecane (8b) was
obtained analogous to 2b from 1.7 g (7.7 mmol) of 8a, 0.55 g (18.5
mmol) formaldehyde (as 30% aqueous solution) and 3.5 g (77.2
mmol) of formic acid. After 24 h at 80Ϫ90°C additional 1.7 g (7.7
mmol) formaldehyde and 3.5 g (76.0 mmol) formic acid were ad-
ded, and the mixture was heated to 80Ϫ90°C for another 24 h. Ϫ
Yield 1.3 g (68%); b.p. 118Ϫ120°C /0.8 Torr. Ϫ C12H26N2O3; mol.
mass 246.35; calcd. C 58.5, H 10.6, N 11.4; found C 58.8, 58.7; H
2.72 (m, 8 H, 4 CH2N), 3.64 (m, 16 H, 2 OCH2).
1,13-Diaza-4,7,10,16-tetraoxacyclooctadeca-2,12-dione (4c) was
synthesized from 1.7 g (16.0 mmol) of 1,5-bis(diamino)-3-oxapen-
tane and 4.0 g (16.0 mmol) of 3,6,9-trioxaundecanedioic acid di-
methyl ester[18d] in 600 ml of dry methanol according to a method
described earlier.[18e] Yield 2.3 g (49%) colourless crystals, m.p.
117Ϫ118°C (acetone). Ϫ C12H22N2O6; mol. mass 290.32; calcd. C
49.6, H 7.6, N 9.7; found C 49.8, 49.8; H 7.7, 7.8; N 9.6, 9.7. Ϫ
1H NMR: δ ϭ 3.49 (m, 8 H, 4 NCH2 ϩ 4 HOCH2), 3.70 (m, 8 H,
4 OCH2 ), 4.02 [s, 4 H, 2 OCH2C(O)], 7.18 (br. s, 2 H, 2 NH ).
1
10.4, 10.3; N 11.5, 11.4. Ϫ H NMR: δ ϭ 2.31 (s, 6 H, 2 NCH3),
2.68 (m, 8 H, 4 CH2N), 3.63 (m, 12 H, 2 OCH2).
1,13-Diaza-4,7,10,-trioxacyclopentadecane-2,12-dione (8c) was
obtained analogous to 4c from 3.5 g (58.8 mmol) ethylenediamine
and 14.7 g (58.8 mmol) of 3,6,9-trioxaundecanedioic acid dimethyl
ester[18d] in 600 ml of dry methanol. The solvent was removed in
vacuo, the residue was purified by column chromatography [Al2O3/
CHCl3, CHCl3-EtOH (20:1)] to yield 5.8 g (40%) ester as colorless
crystals, m.p. 157Ϫ159°C (acetone). Ϫ C10H18N2O5; mol. mass
246.27; calcd. C 48.8; H 7.4; N 11.4; found: C 48.8, 48.7; H 7.5,
1,16-Diaza-4,7,10,13-tetraoxacyclooctaadecane (5a) was obtained
analogous to 4a by reduction of 1.3 g (4.5 mmol) 5c with 2.0 g (52.6
mmol) of lithium aluminium hydride in refluxing tetrahydrofuran
during 24 h. Ϫ Yield 0.7 g (59%); b.p. 125Ϫ125.7°C/0.8 Torr. Ϫ
C12H26N2O4; mol. mass 262.35; calcd. C 54.9, H 10.0, N 10.7;
1
found C 55.0, 54.9; H 10.0, 10.1; N 10.6, 10.7. Ϫ H NMR: δ ϭ
1.95 (s, 2 H, 2 NH ), 2.64 (s, 4 H, NCH2CH2N), 2.68 (m, 4 H, 2
OCH2CH2N), 3.62 (m, 16 H, 8 OCH2).
1
7.4; N 11.1, 11.2. Ϫ H NMR: δ ϭ 3.48 (m, 4 H, 2 NCH2), 3.70
(m, 8 H, 2 OCH2CH2 ), 4.02 (s, 4 H, 2 OCH2C(O)), 7.48 (br. s, 2
H, 2 NH)
N,N-Dimethyl-4,7,10,13-tetraoxa-1,16,-diazacyclooctaadecane
(5b) was obtained analogous to 2b from 2.0 g (7.6 mmol) of 5a,
0.54 g (18.3 mmol) formaldehyde (as 30% aqueous solution) and
3.5 g (76.0 mmol) of formic acid, 80Ϫ90°C. After 24 h additional
0.54 g (18.5 mmol) formaldehyde and 3.5 g (76.0 mmol) formic
acid were added to the reaction mixture, which was heated to
80Ϫ90°C for another 24 h. Ϫ Yield 1.1 g (50%); b.p. 130Ϫ132°C/
0.9 Torr. Ϫ C14H30N2O4; mol. mass 290.41; calcd. C 58.0, H 10.4,
1,4,7,10,13,16,19-Heptaoxacycloheneicosane (9) was a commer-
cial sample.
1,13-Diaza-4,7,10,16,19-pentaoxacycloheneicosane (10) was ob-
tained analogous to 4a by reduction of 1,5 g (4.5 mmol) of 1,13-
diaza-4,7,10,16,19-pentaoxacycloheneicosa-2,12-dione[18e] with 2.0
g (52.6 mmol) of lithium aluminium hydride in boiling tetrahydro-
furan during 24 h.
1
N 9.6; found C 57.8, 58.0; H 10.4, 10.0; N 9.9, 10.0. Ϫ H NMR:
δ ϭ 2.32 (s, 6 H, 2 NCH3), 2.68 (m, 8 H, 4 CH2N), 3.64 (m, 16 H,
2 OCH2).
[1]
Of the Saarbrücken series supramolecular chemistry this is
1,16-Diaza-4,7,10,13-tetraoxacyclooctaadeca-2,15-dione (5c) was
prepared similarly to 4c from 1.1 g (18.0 mmol) ethylenediamine
and 5.4 g (18.0 mmol) of 3,6,9,12-tetraoxatetradecanedioic acid di-
methyl ester[18d] in 600 ml of dry methanol. The solvent was re-
moved in vacuo, the residue was purified by column chromatogra-
phy [Al2O3/CHCl3, CHCl3-EtOH (20:1)] to yield 1.6 g (31%) of
compound 5c, colourless crystals, m.p. 102Ϫ104°C (acetone). Ϫ
C12H22N2O6; mol. mass 290.32; calcd. C 49.6, H 7.6, N 9.6; found
paper 76. Paper 77: Y. Wang, H.-J. Schneider, J. Chem. Soc.,
Perkin Trans. 2, in print.
[2]
For reviews see: Y. Inoue, G. W. Gokel, Cation Binding by
Macrocycles; Marcel Dekker, New York 1990. Ϫ Comprehensive
Supramolecular Chemistry, Vol. 1; G. W. Gokel, Ed., Pergamon/
Elsevier Oxford etc, 1996. Ϫ S. R. Cooper, Crown Compounds:
Towards Future Applications, VCH Publisher, New York,
Weinheim, Cambridge, 1992.
[3]
H.-J. Schneider, V. Rüdiger, O. A. Raevsky, Org.Chem., 1993,
58, 3648.
1
[4]
C 49.8, 48.6; H 7.5, 7.4; N 9.6, 9.7. Ϫ H NMR: δ ϭ 3.46 (m, 4
H, 2 NCH2 ), 3.70 (m, 12 H, 2 OCH2CH2 ), 4.02 [s, 4 H, 2
OCH2C(O)], 7.68 (br. s, 2 H, 2 NϪH).
O. A. Raevsky, V. Ju. GrigorЈev, V. Ju. Kireev, N. S. Zefirov,
Quant. Struct.-Act. Relat., 1992, 11, 49. Ϫ O. A. Raevsky, J.
Phys. Org. Chem., 1997, 10, 405.
[5]
V. P. Solovev, N. N. Strakhova, O. A. Raevsky, V. Rüdiger, H.-
J. Schneider, J.Org.Chem., 1996, 61, 5221.
1,4,7,10,13-Pentaoxacyclododecane (6) was a commercial sample.
[6]
O. A. Raevsky, V. P. Solovev, A. F. Solotnov, H.-J. Schneider, V.
N,N-Dimethyl-4,10,13-tetraoxa-1,7,-diazacyclopentane (7a) was
obtained analogous to 2b from 1.2 g (4.8 mmol) of 7b, 0.35 g (11.7
mmol) formaldehyde (as 30% aqueous solution) and 2.2 g (48.0
mmol) of formic acid, 80Ϫ90°C. Ϫ Yield 0.70 g (53%); b.p.
110Ϫ112°C/0.9 Torr. Ϫ C12H26N2O3 ; mol. mass 246.35; calcd. C
Rüdiger, J.Org.Chem., 1996, 61, 8113.
[7]
R. M. Izatt, K. Pawlak, J. S. Bradshaw, R. L. Bruening, Chem.-
Rev. 1991, 91, 1721. Ϫ R. M. Izatt, J. S. Bradshaw, S. A. Nielsen,
J. D. Lamb, J. J. Christensen, Chem.Rev., 1985, 85, 271. Ϫ J.
J. Christensen, D. J. Eatough, R. M. Izatt, Chem.Rev., 1974,
74, 351.
1388
Eur. J. Org. Chem. 1998, 1379Ϫ1389

Aza Crown Ether Complexes
FULL PAPER
[8]
[9j]
No complexation energies are available in the literature for the
1981, 22, 4665. Ϫ
R. A. Schultz, B. D. White, D. M. Dis-
following ligands: (8,18-dioxo-1,7-N2-18C6); 1,4-N2-18C6, 1,4-
(MeN)218C6, 5,18-dioxo-1,4-N218C6, 2,6-dioxo-1,7-N215C5,
1,4-(MeN)215C5 and 5,15-dioxo-1,4-N215C5. From the ligands
in tables 1 and 2 the complexation of ligand 4a was studied
only with Agϩ in MeCN[8a][8b], Pb2ϩ and Agϩ[8c], Hϩ[8d],with
hong, K. A. Arnold, G. W. Gokel, J. Am. Chem. Soc., 1985,
[9k]
107, 6659. Ϫ
From NMR titration.
[10]
[11]
A. T. Brünger, M. Karplus, Acc.Chem.Res., 1991, 24, 54.
K. Gundertofte, J. Palm, I. Petterson, A. Stamvik, J. Comp.
Chem., 1991, 12, 200.
Tlϩ and Agϩ in MeCN, Me2CO and DMF[8e], with Co2ϩ, Ni2ϩ
,
G. J. Wipff, P. A. Kollmann, Nouv. J. Chim., 1985, 9, 457.
Y. Inoue, Yu. Liu, T. Hakushi in: Cation Binding by Macro-
cycles. Ϫ Y. Inoue, G. W. Gokel, (Eds.) Marcel Dekker, New
York, 1990, 1.
[12]
[13]
Cu2ϩ and Zn2ϩ ions in H2O[8f]; with 4b in CH2Cl2 with chiral
and achiral (phenylalkyl)ammonium thiocyanates[8g]; 7a also
forms complexes with alkylammonium cations in organic sol-
vents[8h][8i] and with metal cations in H2O and CHCl3.[8j]
Thermodynamics of lanthanide complexation in various sol-
vents[8k] and stability constants with Co2ϩ, Ni2ϩ, Cu2ϩ and
[14]
An unpublished X-ray study of analogues 5c of this ligand and
its complexes shows the trans amide conformation in all cases,
but unusual stochiometries and binding modes.
Zn2ϩ ions[8f] were studied with the ligand 8a. Ϫ [8a] V. P. Baran- [15] [15a] V. P. SolovЈev, E. A. Vnuk, N. N. Strakhova, O. A. Raevsky,
nikov, S. S. Guseinov, A. V. Matveev, A. I. VЈyugin, Russ. J.
Thermodynamic of complexation of the macrocyclic polyethers
[8b]
Coord. Chem. (Transl. of Koord. Khim.), 1996, 22, 263. Ϫ
with salts of alkali and alkaline-earth metals. VINITI, Moscow,
[15b]
S. S. Guseinov, V. P. Barannikov, A. I. VЈyagin, G. A. Krestov,
1991, 373pp (Russian). Ϫ
V. P. SolovЈev, L. V. Govorkova,
Zh. Khim. Termodin. Termokhim. (Russ.), 1992, 1, 122. Ϫ [8c] K.
O. A. Raevsky, N. S. Zefirov, Doklady Akademii Nauk (Rus.),
1992, 324, 830.
[16] [16a]
E. Krakowiak, G. E. Maas, J. S. Bradshaw, J. K Hathaway, R.
[8d]
M. Izatt, J. Heterocycl. Chem., 1995, 32, 179. Ϫ
V. P. Vasi-
F. R. Hartley, C. Burgess, R. M. Alcock, Solution Equilib-
[16b]
lЈev, T. D. Orlova, N. Yu. Goncharova, Zh. Fiz. Khim.(Russ.),
ria, Ellis Horwood limited, Chichester, England, 1980. Ϫ
[8e]
1994, 68, 12. Ϫ
A. Jabbari, M. Hasani, M. Shamsipur, J.
M. Beck, I. Nagypal, Chemistry of Complex Equilibria, Akade-
[8f]
Inclusion Phenom. Mol. Recognit. Chem., 1993, 15, 329. Ϫ
miai Kiado, Budapest, 1989. Ϫ [16c] G. A. Seber, Linear Hypoth-
[16d]
E. Luboch, A. Cygan, J. F. Biernat, Inorg. Chim. Acta, 1983,
esis: A General Theory. Hafner, New York, 1966. Ϫ
M.
[8g]
68, 201. Ϫ
L. C. Hodgkinson, M. R. Johnson, S. J. Leigh,
Meloun, J. Havel, Computation of Solution Equilibria. 1. Spec-
trophotometry. University of Purkyne J. E., Brno, 1985.
C. S. Wilcox in Frontiers in Supramolecular Organic Chemistry
and Photochemistry, H.-J. Schneider, H. Duerr (Eds.), VCH,
Weinheim, 1991.
[18] [18a]
N. Spencer, I. O. Sutherland, R. F. Newton, J. Chem. Soc., Per-
[8h]
[17]
kin Trans., 1, 1979, 2193. Ϫ
A. B. Kyte, K. A. Owens, I. O.
Sutherland, R. F. Newton, J. Chem. Soc., Perkin Trans. 1, 1987,
1921. Ϫ [8i] M. R. Johnson, N. F. Jones, I. O. Sutherland, R. F.
[8j]
Newton, J. Chem. Soc., Perkin Trans. 1, 1985, 1637. Ϫ
P.
B. Dietrich, J.-M. Lehn, J. P. Sauvage, Tetrahedron, 1973,
[18b]
Gramain, Y. Frere, Nouv. J. Chim., 1979, 3, 53. Ϫ [8k] F. Arnaud,
Eur. J. Solid State Inorg. Chem., 1991, 28, 229.
28, 1647. Ϫ
G. W. Gokel, B. J. Garcia, Tetrahedron Lett,
J. Chem. Sos., Perk. Trans. 1, 1979, 2193. Ϫ
[18c]
1977, 317. Ϫ
[9] [9a]
[9b]
[18d]
H.-J. Buschmann, Thermochim. Acta, 1986, 102, 179. Ϫ
H.-J. Buschmann, J. Inorg. Chim. Acta, 1986, 125, 31. Ϫ
M. Lehn, P. Vierling, Tetrahedron Lett., 1980, 21, 1323. Ϫ
M. Miyazaki, Y. Shimoishi, H. Miyatama, K. Toeji, J. In-
[9c]
[18e]
J.
org. Chem., 1974, 36, 2033. Ϫ
J. Jurczak, T. Stankiewicz, P.
[9d]
Salanski, S. Kasprzyk, P. Lipkovski, Tetraedron, 1993, 49, 1478.
Ϫ [18f] J. H. Shim, K. B. Chung, M. Tomoi, Bull. Korean Chem.
Soc., 1992, 13, 252. Ϫ [18g] B. Dietrich, J. M. Lehn, J. P. Sauvage,
J. Blanzat, Tetrahedron, 1973, 21, 1629.
[19] [19a]
V. J. Gatto, K. A. Arnold, A. M. Viscariello, S. R. Miller, C. R.
[9e]
Morgan, G. W. Gokel, J. Org. Chem., 1986, 51, 5373. Ϫ
R.
B. Davidson, R. M. Izatt, J. J. Christensen, R. A. Schultz, D.
[f]
M. Dishong, G. W. Gokel, J. Org. Chem., 1984, 49, 5080. Ϫ
R. M. Izatt, R. E. Terry, B. L. Haymore, L. D. Hansen, N.
K. Dalley, A. G. Avondet, J. J. Christensen, J. Am. Chem. Soc.,
1976, 98, 7620. Ϫ [19b] R. M. Izatt, D. V. Dearden, P. R. Brown,
J. S. Bradshaw, J. D. Lamb, J. J. Christensen, ibid., 1983, 105,
1785.
B. G. Cox, H. Schneider, Pure Appl. Chem., 1989, 61, 171. Ϫ
[9g]
Y. Inoue, Y. Liu, L.-H. Tong, M. Ouchi, T. Hakushi,
J.Chem.Soc., Perkin Trans. 2, 1993, 1947. Ϫ [9h] J. J. Christensen,
D. J. Eatough, R. M. Izatt, Chem.Rev., 1974, 74, 351. Ϫ H. K.
[9i]
Frensdorf, J. Am. Chem. Soc., 1971, 93, 600. Ϫ
A. Masuy-
[97383]
ama, Y. Nakatsuji, I. Ikeda, M. Okahara, Tetrahedron Lett.,
Eur. J. Org. Chem. 1998, 1379Ϫ1389
1389