Structural study of (±) alkyl 3-hydroxy-1-azabicyclo [2.2.2] octane-3-carboxylates

Article abstract of DOI:10.1016/S0022-2860(02)00490-8

A series of α-hydroxyesters derived from (±) 3-hydroxy-l-azabicyclo[2.2.2]octane-3-carboxylic acid was synthesised and studied by IR and NMR spectroscopy. The combined use of ¹H-¹H COSY and ¹H-¹³C correlation spectra of these compounds helped in the unambiguous assignments of the bicyclic carbon and proton resonances. The crystal structure of ethyl (±) 3-hydroxy-1-azabicyclo [2.2.2] octane-3-carboxylate was determined by X-ray diffraction.

Full text of DOI:10.1016/S0022-2860(02)00490-8

Journal of Molecular Structure 644 (2003) 171–179
www.elsevier.com/locate/molstruc
Structural study of (^) alkyl 3-hydroxy-1-
azabicyclo[2.2.2]octane-3-carboxylates
a,
a
a
b
b
M.S. Arias-Perez , A. Cosme , E. Galvez , J. Sanz-Aparicio , I. Fonseca , J. Bellanato
c
*
´
´
a
´ ´ ´ ´
Departamento de Quımica Organica, Universidad de Alcala, Alcala de Henares, 28871 Madrid, Spain
b
´
Departamento de Cristalografıa, Instituto Rocasolano, CSIC, Serrano 119, 28006 Madrid, Spain
cInstituto de Optica, CSIC, Serrano 119, 28006 Madrid, Spain
Received 17 July 2002; revised 20 September 2002; accepted 20 September 2002
Abstract
A series of a-hydroxyesters derived from (^) 3-hydroxy-1-azabicyclo[2.2.2]octane-3-carboxylic acid was synthesised and
studied by IR and NMR spectroscopy. The combined use of 1H–1H COSY and 1H–13C correlation spectra of these compounds
helped in the unambiguous assignments of the bicyclic carbon and proton resonances. The crystal structure of ethyl (^) 3-
hydroxy-1-azabicyclo[2.2.2]octane-3-carboxylate was determined by X-ray diffraction.
q 2002 Elsevier Science B.V. All rights reserved.
Keywords: 3-Hydroxy-1-azabicyclo[2.2.2]octane-3-carboxylates; a-Hydroxyesters; Quinuclidine derivatives; NMR spectroscopy; X-ray
crystallography
1. Introduction
which exhibit some apparent chemical analogies and a
variety of common pharmacological effects [1–5].
One approach towards the development of more
receptor-selective ligands is the incorporation of
conformational restrictions, due to the critical
influence of the stereochemical effects on the
ligand–receptor interactions. Thus, several of the
structure–activity relationships developed have been
rationalised in terms of the ability of the low-energy
conformations to dock optimally on the pharmaco-
phore framework.
There has been considerable interest in the study of
the molecular mechanisms operating at the different
families of neurotransmitters receptors in the central
nervous system. Among them, the nicotinic cholin-
ergic [1–3], serotonergic [1,3,4], glycinergic [1] and
GABAergic receptors [1,5] (ligand-gated ion channel
receptors, LGICR) are important due to their roles in
several cognitive disorders such as Alzheimer’s and
Parkinson’s diseases. A number of agonists and
antagonists of these receptors has been described
As a part of a research program aimed at the
development of new ligands for ionotropic recep-
tors [6], we are currently involved in studies in
which the 1-azabicyclo[2.2.2]octane (quinuclidine)
system is being utilised as a conformationally
*
Corresponding author. Tel.: þ34-91-885-4699; fax: þ34-91-
885-4686.
´
E-mail address: selma.arias@uah.es (M.S. Arias-Perez).
0022-2860/03/$ - see front matter q 2002 Elsevier Science B.V. All rights reserved.
PII: S0022-2860(02)00490-8

´
M.S. Arias-Perez et al. / Journal of Molecular Structure 644 (2003) 171–179
172
Scheme 1.
restricted framework [7], bearing in mind that
several quinuclidine derivatives have been ident-
ified as potent antagonists and agonists of muscar-
inic [8,9] and 5-HT3 receptors [3,4]. We report
here the structural study of some a-hydroxyesters,
3–6, derived from (^) 3-hydroxy-1-azabicy-
clo[2.2.2]octane-3-carboxylic acid (Scheme 1), by
IR and NMR spectroscopy, and the crystal structure
of (^) ethyl 3-hydroxy-1-azabicyclo[2.2.2]octane-
3-carboxylate, 3. The structural features of these
compounds are as follows: (a) the conformation of
the ethanolamine chain is constraint by the
quinuclidine ring as in the potent muscarinic
antagonists 3-quinuclidinyl benzilate (QNB) and
3-quinuclidinyl atrolactate (QNA) [8–10]; (b) the
hydroxy and carboxylate groups are attached to the
position 3 of the quinuclidine system leading to a
different spatial location of both groups from that
in QNB and QNA; (c) the greater conformational
freedom correspond to the alkoxy side chain where
an aryl group was introduced at positions b or g.
These restricted derivatives are analogues of the
nipecotinic acid that has been shown to be a potent
inhibitor of the GABA-uptake process [11]. Fur-
thermore, a fixed conformation of the flexible
b-alanine, a ligand of the glycinergic receptor
with dual behaviour [11], can be mimicked in 3–6.
marised in Table 1 together with the refinement
procedures [12–16]. The titled compound crystal-
lisation was very poor and only a small crystal was
suitable for crystallographic analysis. The crystal
structure could be solved, even H atoms were found,
but a good refinement was not achieved.
The IR spectra of 3 were recorded on a Perkin–
Elmer 599B spectrophotometer in the solid state
(KBr) and in CDCl3 solution (0.06 M) using
0.2 mm NaCl cells. Spectra of very dilute CCl4
solutions were taken using 4 cm quartz cells. The
reported wavenumbers were estimated to be
accurate to within ^4 cm21
.
All NMR spectra (1H, 13C, double resonance-
decoupling-experiments, DEPT, 1H–1H COSY-45
1
and H–13C HETCOR) were recorded on a Varian
UNITY-300 spectrometer in CDCl3 at 298 K using
standard pulse sequences. Lorentz–Gauss trans-
formation (LB ¼ 20.8; GF ¼ 0.6 and GFS ¼ 0.2)
1
was used to improve the resolution of the H-NMR
spectra [17]. In the case of 3, the agreement
between the observed spectrum and the parameters
established from analysis with the LAOCOON III
program [18] was verified by simulation of the
calculated spectrum.
2.1. Synthesis
The a-hydroxyesters 3–6 were synthesised from
the cyanohydrin 1 and the appropriate alcohol via
the iminoester hydrochloride 2 (Scheme 1), by an
improved modification of the method previously
described for 3 [19]. Thus, 3 was obtained in a
2. Experimental
The main crystallographic data and the structure
determination conditions for compound 3 are sum-

´
M.S. Arias-Perez et al. / Journal of Molecular Structure 644 (2003) 171–179
173
Table 1
3. Results and discussion
Experimental data and structure refinement procedures for 3
3.1. Crystal structure of 3
Crystal data
Formula
C10H17NO3
The PLUTO view of the molecule is displayed
in Fig. 1, which also shows the atomic numbering.
The atomic coordinates for all non-hydrogen atoms
are given in Table 2 and some selected bond
lengths, bond and torsion angles are collected in
Tables 3 and 4.
Crystal size (mm)
Symmetry
0.02 £ 0.05 £ 0.15
Monoclinic, P21=n
Unit cell determination Least-squares fit from 35 reflexions
(u , 198)
˚
Unit cell dimensions
6.481 (2), 16.298 (2), 10.106 (2) A
90.0, 97.57 (2), 90.08
1058.2 (4), 4
3
Packing: V (A ), Z
˚
All bond distances and valence angles corre-
spond to the expected values. As can be see in
Fig. 1, the bicyclic system shows the usual skew
conformation, though torsions are smaller than
those found in the (^) 3-(4-methylphenyl)-1-
azabicyclo[2.2.2]octan-3-ol [20], which presents
the same substitution pattern, but the ethoxycarbo-
nyl group is replaced by an aryl moiety. Thus, the
C2–N· · ·C4–C3, C6–N· · ·C4–C8 and C6–
N· · ·C4–C5 torsion values are 4.5, 2.6 and 1.28,
respectively, for 3 and 10.9, 9.4 and 9.48 for the 3-
aryl-3-quinuclidinol. However, the poor refinement
attained does not allow making any conclusion.
The biggest torsion is at the disubstituted carbon
C3.
Dc (g/cm23), M, F(000) 1.251, 199.249, 432
m (Mo Ka) (cm21
)
0.859
Experimental data
Technique
Four circle difractometer: Enraf Nonius
CAD-4; bisecting geometry; graphite
oriented monochromator, Mo Ka; v/2u
scans; detector apertures 1 £ 18, up to
umax 308, 1.5 min/refex
Number of reflexions
Measured
3178
Independent
Observed
3053
719 (2sðIÞ criterion)
28 8, 0 22, 0 14 (sin u/l )max 0.7
0.007
Range of hkl
Value of Rint
Solution and refinement
Solution
Direct methods
Refinement
H atoms
Least-squares on Fobs with 1 blocks
Difference synthesis
Fig. 2 shows a schematic view of the molecular
packingintheunitcell. Themoleculesareheldtogether
by hydrogen bonds between the hydroxy group (O3 as
donor) and the bicyclic N atom: O3· · ·N (x 2 1/2,
2y þ 1/2, z 2 1/2) ¼ 2.79, O3–H03 ¼ 0.99y, HO3–
W-scheme
Empirical as to give no trends in
kwD2Fl versus klFoll or ksin u/ll
0.02
Final shift/error
Final DF peaks
Final R and Rw
23
˚
0.21 e/A
˚
0.091, 0.089
N ¼ 1.87 A and O3–H03· · ·N ¼ 1528.
Computer and programs VAX 11/750, MULTAN 80 [12], X-
RAY76 [13], PESOS [14], PARST [15]
Pharmacologically interesting are the intramole-
˚
cular distances N1· · ·O3, 3.42 A, similar to that
Scattering factors
International tables for X-ray crystallo-
graphy [16]
˚
reported for the 3-quinuclidinol (,3.5 A) [10] and
N1· · ·O2, close to that found for the selective GABAB
receptor agonist baclofen in solid state (baclofen HCl)
[21,22]. Moreover, in 3 the spatial arrangement of the
structural fragment C9–C3–C4–C8 is similar to the
related Ca–Cb and Cb–Cg bonds in the baclofen
HCl and shows a reasonably good concordance with
one of the bioactive conformations proposed for the
GABA molecule [6].
Anomalous dispersion
International tables for X-ray crystallo-
graphy [16]
yield of 91% (versus 55% [19]) by heating the
aqueous solution of the corresponding iminoester
hydrochloride at 40–50 8C for 90 min. In the case
of 5, the formation of 2 was carried out using 1,4-
dioxane as solvent. As the size of the alcohol
increased, the yield of the a-hydroxyester
decreased: 84% (4), 64% (5) and 52% (6).
3.2. Infrared spectra
In the OH stretching region, the infrared spectrum
of 3 in the solid state presented a strong complex
double absorption between 3500 and 2300 cm21
,

´
M.S. Arias-Perez et al. / Journal of Molecular Structure 644 (2003) 171–179
174
Fig. 1. PLUTO view of 3 showing the atomic numbering.
centred at about 2850 cm21. Following Claydon and
Sheppard [23], the two absorptions can be explained
as originated by Fermi resonance of nOH with
overtones and/or combination frequency modes. As
it could be expected, X-ray data revealed the existence
of a strong intermolecular hydrogen bond between the
OH group and the basic nitrogen atom of the bicyclic
system. This behaviour is quite similar to that reported
previously for alkyl 3-hydroxytropan-3-carboxylates
which also showed a strongly hydrogen bonded
system [24].
indicates the presence of some intermolecular O–
H· · ·N bonding.
In the carbonyl region, the spectrum registered in
the solid state presented a very strong band at
1724 cm21 which is assigned to the stretching
Table 2
Atomic coordinates and thermal parameters for non-hydrogen
atoms of 3
a
Atom
x
y
z
Ueq
Upon dilution in CCl4 (0.001 M), the absorption
due to strong hydrogen bonding disappeared and a
sharp band appeared at 3598 cm21. This band is
attributed to free OH group and was accompanied by a
broad band at 3540 cm21, which is tentatively
assigned to a weak intramolecular hydrogen bond
between the OH and CyO groups, forming a five
member ring. In CDCl3 solution (0.06 M), a double
band was also observed with maxima at 3578 and
3545 cm21 which are interpreted in terms of solvent-
bonded and intramolecularly bonded OH groups,
respectively. Therefore, the infrared results also
suggest the existence of intramolecular bonding in
CDCl3 solution. Moreover, at the concentration used,
a weak broad absorption centred at 2950 cm21
C3
C2
N
0.010 (2)
0.195 (2)
0.122 (3)
0.303 (5)
0.291 (5)
0.286 (4)
0.221 (4)
0.243 (3)
0.318 (5)
0.385 (4)
0.364 (3)
0.376 (5)
0.420 (3)
0.487 (4)
0.522 (3)
0.392 (4)
0.232 (3)
0.302 (6)
0.418 (5)
0.549 (4)
0.550 (2)
0.446 (5)
0.370 (4)
0.471 (5)
0.582 (4)
0.211 (4)
0.244 (5)
0.157 (4)
0.205 (3)
0.118 (5)
0.222 (5)
3 (3)
5 (4)
3 (2)
5 (4)
5 (4)
3 (3)
7 (4)
8 (4)
3 (3)
5 (2)
7 (4)
7 (6)
9 (3)
5 (2)
C7
C8
C4
C5
C6
C9
O1
C10
C11
O2
O3
20.041 (2)
20.229 (4)
20.184 (2)
20.151 (4)
0.034 (3)
0.036 (2)
0.208 (3)
0.238 (3)
0.451 (4)
20.091 (4)
20.010 (4)
For atomic labelling, see Fig. 1; e.s.d. values are given in
parentheses.
P P
i
a
Ueq ¼ 1=3
½Uijapi ajpaiaj cosðai; ajފ £ 103:
j

´
M.S. Arias-Perez et al. / Journal of Molecular Structure 644 (2003) 171–179
175
Table 3
˚
Bond lengths (A) and angles (8) in 3
of the quinuclidine system, its proton magnetic
parameters have often not been reported in sufficient
detail and, to our knowledge, have not been
previously described for the derivatives 3–6.
Bond lengths
˚
Length (A)
˚
Length (A)
The 13C NMR chemical shifts are tabulated, with
the signal assignments, in Table 5. Substituent steric
and electronic effects on the 13C chemical shifts [25,
26] were taken into consideration. The distinction
between the C-5 and C-8 resonances was based on
the different upfield-shifting g-gauche effects exerted
by the hydroxy and alkoxycarbonyl groups [25,26].
The value of the g-gauche effect of the OH group was
deduced using 7 as model and is similar—ca. 6 ppm—
to that estimated for this compound in (CD3)2SO [7]
and other quinuclidine [20] and bicyclic derivatives
[27]. Taking into account that a smaller value has
been reported for the g-gauche effect of an alkox-
ycarbonyl moiety [25,26], the signal at highest-field
(20 ppm) must correspond to C-8 (Scheme 1). Using
compounds 3–7 as models a mean value of around
3 ppm can be proposed for the g-gauche effect exerted
by the COOR group on the bicyclic carbons of the
quinuclidine system.
Bond
Bond
C2–C3
C3–C9
N–C2
1.58 (5)
1.52 (9)
1.47 (6)
1.45 (8)
1.50 (8)
1.57 (5)
1.20 (5)
1.51 (4)
C3–C4
C3–O3
N–C7
1.52 (4)
1.41 (9)
1.50 (7)
1.54 (4)
1.49 (9)
1.33 (5)
1.43 (7)
N–C6
C7–C8
C4–C5
C9–O1
O1–C10
C4–C8
C5–C6
C9–O2
C10–C11
Bond angles (8)
Angle
Value (8)
107.3 (4)
108.3 (3)
114.5 (2)
111.4 (2)
111.0 (3)
109.1 (2)
109.5 (3)
111.8 (2)
108.0 (4)
115.5 (3)
116.3 (2)
Angle
Value (8)
111.6 (2)
108.7 (3)
106.6 (2)
110.7 (4)
108.6 (2)
110.3 (2)
105.2 (3)
111.0 (4)
122.2 (3)
122.3 (4)
106.5 (1)
C9–C3–O3
C4–C3–C9
C2–C3–C9
C3–C2–N
C2–N–C7
N–C7–C8
C3–C4–C8
C3–C4–C5
N–C6–C5
C3–C9–O1
C9–O1–C10
C4–C3–O3
C2–C3–O3
C2–C3–C4
C2–N–C6
C6–N–C7
C7–C8–C4
C5–C4–C8
C4–C5–C6
C3–C9–O2
O1–C9–O2
O1–C10–C11
At 300 MHz, the 1H NMR spectra of 3–7 are very
similar. The interpretation of the 1H–1H COSY
spectra was based on the unambiguous assignment
For atomic labelling, see Fig. 1; e.s.d. values are given in
parentheses.
vibration of free CyO groups, in agreement with
X-ray data and the results obtained for the above
mentioned alkyl 3-hydroxytropan-3-carboxylates
[24]. However, in CCl4 solution, in accordance with
results in the OH stretching region, two bands
appeared at 1743 and 1720 cm21, which are assigned
to free and intramolecularly bonded carbonyl groups.
In CDCl3 solution, the higher frequency band was
only observed as a shoulder of the predominant band
Table 4
Selected torsion angles (8) in 3
Angle
Value (8)
C4–C5–C6–N
N–C2–C3–C4
N–C2–C3–C9
N–C7–C8–C4
C9–C3–C4–C8
C9–C3–C4–C5
C3–C4–C5–C6
C7–C8–C4–C3
C2–C3–C9–O1
C2–C3–C9–O2
C4–C3–C9–O1
C4–C3–C9–O2
O3–C3–C9–O1
O3–C3–C9–O2
O2–C9–O1–C10
C3–C9–O1–C10
C9–O1–C10–C11
2 (5)
7 (6)
127 (5)
4 (5)
173 (4)
271 (5)
259 (5)
58 (5)
at 1715 cm21
.
3.3. NMR spectra
22 (6)
179 (5)
116 (5)
263 (7)
2123 (5)
58 (7)
The a-hydroxyesters 3–6 and the parent (^) 1-
azabicyclo[2.2.2]octan-3-ol, 7, were studied in depth
by 1H and 13C NMR in CDCl3. The assignment of the
bicyclic proton and carbon resonances was achieved
23 (7)
178 (4)
2174 (4)
1
1
by the combined use of H–1H COSY and H–13C
correlation spectra, homonuclear spin decoupling
experiments and our previous studies on related
quinuclidine derivatives [7,20]. Due to the complexity
For atomic labelling, see Fig. 1; e.s.d. values are given in
parentheses.

´
M.S. Arias-Perez et al. / Journal of Molecular Structure 644 (2003) 171–179
176
Fig. 2. View of the molecular packing in the unit cell of 3.
of the signals for H-3 (7), H-4 and C-2 protons, owing
to their shape and chemical shifts. These signals are
well differentiated, while the other bicyclic proton
resonances exhibit higher complexity and appear
partially overlapped. Thus, for 3, the cross-peaks
observed between the apparent quintet due to H-4 at
1.90 ppm and the high-field signals at ca. 2.07, 1.53
and 1.32 ppm indicated that these multiplets must
correspond to C-5 and C-8 protons. Therefore, the
signals contained in the low-field region must be due
to C-2, C-6 and C-7 protons, in agreement with the
deshielding effect of the nitrogen atom.
The differentiation between the signals of C-5 and
C-8 protons could be made from the heteronuclear
1H–13C correlated spectra, once the resonances of the
respective carbons were known. In the case of 3, the
observation of the following correlations
d ¼ 20.3/2.07 and 1.32 ppm and d ¼ 22.9/1.53 ppm
allows the assignment of the multiplets at 2.07 and
1.32 ppm to the C-8 protons, while the C-5 proton
resonances must be partially overlapped at ca.
1.53 ppm. The C-6 and C-7 protons were assigned
on the basis of their correlations with the C-5 and C-8
protons. From these statements, the assignment of the
diastereotopic protons of each methylene group was
based on the analysis of the COSY cross-peaks due to
W long-range couplings (4J) and the coupling
modifications observed in the different irradiation
experiments. Only the ill-resolved multiplet corre-
sponding to C-6 protons could not be analysed.
Finally, the analysis of the 1H–13C correlated
spectrum allowed the distinction between the chemi-
cal shifts of C-6 and C-7, once the resonances of the
respective protons were established.
Table 5
13C chemical shifts for the a-hydroxyesters 3–6 and (^) 1-
azabicyclo[2.2.2] octan-3-ol (7)
Chemical shifts d (ppm)
3
4
5
6
7
C-8
20.3
22.9
31.5
45.9
46.5
58.3
74.0
20.1
22.7
31.0
45.6
46.2
58.1
73.8
20.3
22.8
31.6
45.9
46.5
58.2
74.3
20.1 18.8
22.8 24.7
31.2 28.3
45.6 46.2
46.2 47.2
58.0 57.9
73.8 67.0
C-5
C-4
C-6
C-7
C-2
C-3
C-9
175.2 175.1 175.2 175.0
C-10
C-11
C-12
C-14 (18)
C-15 (17)
C-16
C-13
61.8
14.2
65.7
34.9
65.8
34.3
64.9
30.0
32.0
The proton chemical shifts of compounds 3–7 are
shown in Table 6. For 3, the multiplets due to H-71, C-5
and C-8 protons were considered as parts of
the respective spin systems and analysed with the
LAOCOON III program [18]. The observed 1H–1H
coupling constants for the quinuclidine moiety are very
128.4a 130.0 128.2a
128.6a 128.6 128.3a
126.5 132.6 126.0
137.2 135.8 140.6
a
Interchangeable assignment.

´
M.S. Arias-Perez et al. / Journal of Molecular Structure 644 (2003) 171–179
177
Table 6
1H chemical shifts for the a-hydroxyesters 3–6 and 7 in CDCl3
Chemical shits d (ppm)
3
4
5
6
7a
H-21 (dd)
H-22 (d)
H-71
3.51
3.45
3.44
3.52
2.99 (m)
2.47 (m)
2.80
2.71 (dd)
2.88b
2.72
2.71
2.76
2.87
2.86
2.91
H-72
2.82
2.81
2.78
2.87–2.74
2.87–2.74
2.87–2.74
2.10
2.70–2.59
2.54
H-61
2.77–2.67
2.77–267
2.07b
2.78–2.68
2.78–2.68
2.05
2.78–2.69
2.78–2.69
2.04
H-62
2.70–2.59
1.87
H-81
H-4
1.90
1.57b
1.82
1.81
1.90
1.68
H-51
1.44–1.37
1.44–1.37
1.29
1.48–1.34
1.48–1.34
1.30
1.67–1.49
1.67–1.49
1.36
1.38
H-52
1.50b
1.59
H-82
1.32b
1.26
H-10
4.22 (q)
4.46 (m)c
4.39 (m)c
2.99 (t)c
4.42 (m)c
4.34 (m)c
2.95 (t)c
4.24 (m)d
4.20 (m)d
2.02 (tt)d
2.71 (t)d
7.21 (dd)e
7.29 (dd)e
7.30 (tt)e
3.00
H-11
1.25 (t)
H-12
H-14 (18)
H-15 (17)
H-16
7.21 (dd)e
7.28 (dd)e
7.30 (tt)e
3.10
7.13 (m)f
7.26 (m)f
OH (s br)
3.88
3.20
4.77
Abbreviations: b, broad; d, doublet; dd, doublet of doublets; m, multiplet; q, quartet; s, singlet; t, triplet; tt, triplet of triplets.
4
a
H-3 (m): 3.69 ppm, 3JH21–H3 ¼ 8.3, 3JH22–H3 ¼ 3.6, 3JH3–H4 ¼ 3.4, l JH3–H82l ¼ 1.3 Hz.
b
Values deduced from the analysis of the respective spin subsystems with the LAOCOON III program: ABCMX (H-71, H-72, H-61, H-81 and
H-82); ABMPXZ (H-71, H-72, H-81, H-4, H-51 and H-82); ABMPXY (H-61, H-62, H-81, H-4, H-51 and H-52). Error ^0.01 ppm.
These protons appear as four-spin ABX2 systems: 2JH101–H102 ¼ 210.9, 3JH10–H11 ¼ 6.8 (4) and 6.7 Hz (5).
c
d
e
3
These protons appear as an ABM2X2 system: 2JH101–H102 ¼ 210.9, 3JH10–H11 ¼ 6.5 and JH11–H12 ¼ 7.7 Hz.
4
4
3JH14–H15 ¼ 3JH17–H18 ¼ 7.3 (4), 7.2 (6); 3JH15–H16 ¼ 3JH16–H17 ¼ 7.5 (4), 7.1 (6); l JH14–H16l ¼ l JH16–H18l ¼ 1.7 (4),
1.6 Hz (6).
f
4
4
These protons absorb as a four-spin AA0MM0 system: 3JH14–H15 ¼ 3JH17–H18 ¼ 8.3; l JH14–H18l ¼ l JH15–H17l ¼ 2.3; 5JH14–
H17 ¼ 5JH15–H18 ¼ 0 Hz.
similar to those described for related compounds [7,20,
28]. The values deduced for 3 are given in Table 7.
toxy-1-azabicyclo[2.2.2]octane-3-carboxylic acid
hydrochloride [7]; in this case, a slightly distorted
CyO/C2 eclipsed conformation (torsion angle C2–
C3–C9–O ¼ 2198) is preferred. Similar trends have
been reported for simple acyclic acids and esters [29],
suggesting that the quinuclidine system does not
influence significantly the spatial dispositions of the
COOEt and COOH groups. However, the confor-
mation of the C–C(yO) (C3–C9) system in these
derivatives seems to be particularly likely to be
governed by the different nature of both the group
bonded to the dicoordinate oxygen atom (Et versus H)
and the other substituent attached to the same bicyclic
carbon (OH versus OCOR). Consequently, their
behaviour may be attributed to a combination of
steric and charge interaction factors [29]. The
conformation adopted in the crystal structures may
be related mainly with the different hydrogen-bonding
3.4. Conformational study
Concerning the conformational preferences of the
substituents attached to the 3-position, the X-ray data
indicated that the ethoxycarbonyl moiety in 3 displays
in the solid state a conformation in which the
dicoordinate oxygen atom (O1) is practically eclipsed
with the bicyclic carbon C2 (Fig. 1, torsion angle
C2–C3–C9–O1 ¼ 228). The carbonyl group adopts
an almost ideal staggered (bisecting) arrangement
with respect to C4 and the oxygen atom of the
hydroxy group (O3), with torsion angles: C4–C3–
C9–O2 ¼ 2638 and O3–C3–C9–O2 ¼ 588. These
results are in sharp contrast with those found for the
carboxylic group in the related (^) 3-diphenylace-

´
M.S. Arias-Perez et al. / Journal of Molecular Structure 644 (2003) 171–179
178
Table 7
1H–1H coupling constants for the a-hydroxyester 3
additionally stabilised by weak intramolecular hydro-
gen bonding with participation of the carbonyl oxygen
atom (O2) as hydrogen-bond acceptor, in accordance
with the IR results. Furthermore, this conformation
only involves a small distortion of the geometry
observed for 3 by X-ray diffraction.
In summary, it may be proposed that the quinucli-
dine system does not exert a significant influence on
the conformational preferences of the alkoxycarbonyl
moiety in the a-hydroxyesters 3–6. In the solid state,
the preferred conformation was found to be an almost
ideal staggered form, in which the carbonyl group
adopts a bisecting geometry with respect to the
hydroxy group and the bicyclic carbon C4. NMR
and IR data account for a conformational heterogen-
eity with a significant contribution of an almost
CyO/OH eclipsed form, stabilised by a weak
intramolecular hydrogen bond, in CDCl3 solution.
3J (Hz)
2J (Hz)
H4–H51
3.01
3.33
3.25
2.74
10.61
5.51
5.49
10.48
10.39
5.70
4.59
10.58
7.2
H21–H22a
H51–H52
H71–H72
H81–H82
214.2
H4–H52
213.10
213.27
213.10
H4–H81
H4–H82
4
H51–H61
H51–H62
H52–H61
H52–H62
H71–H81
H71–H82
H72–H81
H72–H82
H10–H11a
l Jl (Hz)
H21–H72a
H22–H62a
H51–H81
H61–H71
2.1
2.0
2.90
2.35
Values calculated from the analysis of the respective spin
subsystems by means of the LAOCOON III program; error ^0.05 Hz.
a
Values deduced from the analysis of the spectrum. The values of
these coupling constants are 2JH21–H22 ¼ 214.3 (4–6) and
4
213.9 (7); l JH21–H72l ¼ 2.1 (4 and 5), 2.2 (6 and 7).
Acknowledgements
network. With regard the alkoxy C–O bond, the
preference is for a planar Z conformation (O2–C9–
O1–C10 ¼ 238) as was expected [29].
´
This work was supported by the Spanish Comision
´
Interministerial de Ciencia y Tecnologıa (Project
´
2FD97-0388-C02-02) and the University of Alcala.
1
From the H and 13C parameters, which show a
great similarity, it can be assumed that in CDCl3
solution, the hydroxy and alkoxycarbonyl groups
exhibit the same behaviour. However, these data give
little information about the conformational prefer-
ences of these groups, which arises from the proton
chemical shifts. By comparing the a-hydroxyesters
3–6 and the parent 3-quinuclidinol 7, it was found
that the introduction of an alkoxycarbonyl group at the
3-position exerts a small deshielding effect on the H-
4, H-51 and C-2 protons (Table 6). The effect is
slightly greater for H-21, in a cis orientation
(Scheme 1) and the value of Dd H21–H22 ¼ 0.7–
0.8 ppm in derivatives 3–6 is near to that observed for
7 (0.5 ppm). These facts can tentatively be attributed
to a high flexibility of the alkoxycarbonyl moiety in
solution, suggesting a conformational equilibrium
between eclipsed and bisecting conformations by
rotation around the C–C(yO) (C3–C9) bond.
Although any preference can be unequivocally
asserted, it is reasonable to suppose that an almost
eclipsed orientation of the carbonyl group with
respect to O3 (OH group) should make a significant
contribution. This spatial disposition should be
References
´
[1] M.H. Aprison, E. Galvez, K.B. Lipkowitz, J. Neurosci. Res. 43
(1996) 127.
[2] R.A. Glennon, M. Dukat, Pharm. Acta Helv. 74 (2000) 103.
´
[3] A. Morreale, I. Iriepa, E. Galvez, Curr. Med. Chem. 9(2002)99.
[4] F.D. King, J.B. Jones, G.J. Sanger, 5-Hydroxytryptamine-3-
Receptor Antagonists, CRC Press, Boca Raton, FL, 1994.
[5] P. Krogsgaard-Larsen, B. Frlund, U. Kristiansen, K. Fryden-
vang, B. Ebert, Eur. J. Pharm. Sci. 2 (1997) 355.
´
[6] E. Galvez-Ruano, I. Iriepa, A. Morreale, D.B. Boyd, J. Mol.
Graphics Modell. 20 (2001) 183 and references cited therein.
´
´
[7] M.S. Arias-Perez, A. Cosme, E. Galvez, M.J. Santos, M.
Martinez-Ripoll, E. Matesanz, J. Mol. Struct. 565/566 (2001)
353.
[8] J. Saunders, M. Cassidy, S.B. Freedman, E.A. Harley, L.L.
Iversen, C. Kneen, A.M. McLeod, K.J. Merchant, R.J. Snow,
R. Baker, J. Med. Chem. 33 (1990) 1128.
[9] V.I. Cohen, R.E. Gibson, L.H. Fan, R. De la Cruz, M.S. Gitler,
E. Hariman, R.C. Reba, J. Med. Chem. 34 (1991) 2989.
[10] F.I. Carrol, P. Abraham, K. Parham, R.C. Griffith, A. Ahmad,
M.M. Richard, F.N. Padilla, J.M. Witkin, P.K. Chiang, J. Med.
Chem. 30 (1987) 805.
[11] V. Schmieden, H. Betz, Mol. Pharmacol. 48 (1995) 919.

´
M.S. Arias-Perez et al. / Journal of Molecular Structure 644 (2003) 171–179
179
[12] P. Main, S.J. Fiske, S.E. Hull, L. Lessinger, G. Germain, J.-P.
Declercq, M.H. Woolfson, MULTAN 80: A System of
Computer Programs for Automatic Solution of Crystal
Structures from X-ray Diffraction Data, University of York,
England and Louvain, Belgium, 1980.
[21] C. Chang, D.S.C. Yang, C.S. Yoo, B. Wang, J. Pletcher, M.
Sax, C.F. Terrence, Acta Cryst. B38 (1982) 2065.
[22] V.N. Goka, G. Schlewer, J.M. Linget, J.P. Chambon, C.G.
Wermuth, J. Med. Chem. 34 (1991) 2547.
[23] M.F. Claydon, N. Sheppard, Chem. Commun. D23 (1969)
359.
´
[24] C. Burgos, M.L. Izquierdo, M.S. Arias, E. Galvez, E.
[13] J.M. Stewart, P.A. Manchin, C.L. Dickinson, H.L. Ammon,
H. Heck, H. Flack, The X-ray 76 System, Technical Report
TR-446, Computer Science Center, University of Maryland,
College Park, MD, 1976.
Matesanz, M. Martinez-Ripoll, J. Sanz-Aparicio, I. Fonseca,
´
J. Bellanato, P. Lopez-Ruiz, J. Mol. Struct. 301 (1993) 95.
[14] M. Martinez-Ripoll, F.H. Cano, PESOS: A Computer Program
for the Automatic Treatment of Weighting Schemes, Institute
Rocasolano, CSIC, Madrid, Spain, 1975.
[25] E. Breitmaier, W. Voelter, Carbon-13 NMR Spectroscopy,
High-Resolution Methods and Applications in Organic
Chemistry and Biochemistry, Verlag Chemie, Weinheim,
1987.
[15] M. Nardelli, Comput. Chem. 7 (1983) 95.
[16] International Tables for X-Ray Crystallography, vol. IV,
Kyonch Press, Brmingham, England, 1974.
[26] F.W. Wehrli, A.P. Marchand, S. Wehrli, Interpretation of
Carbon-13 Spectra, Second ed., Wiley, Chichester, 1988.
´
[27] M.S. Arias-Perez, A. Alejo, A. Maroto, Tetrahedron 53 (1997)
[17] H. Gu¨nther, NMR Spectroscopy, Wiley, Chichester, 1995.
[18] S. Castellano, A.A. Bothner-By,J. Chem.Phys. 41(1963)3863.
´ ´
[19] G.G. Trigo, M. Martınez, E. Galvez, R. Cabezas, J. Heterocycl.
13099.
[28] A.P. Marchand, Stereochemical Applications of NMR Studies
in Rigid Bicyclic Systems, Verlag Chemie, Weinheim, 1982.
[29] E.L. Eliel, S.H. Wilen, N. Mander, Stereochemistry of Organic
Compounds, Wiley, New York, 1994.
Chem. 18 (1981) 1507.
´
[20] M.S. Arias, A. Cosme, E. Galvez, F. Florencio, I. Fonseca, J.
Sanz, Magn. Reson. Chem. 29 (1991) 1130.