Chiral bicyclo[3.3.1]-3,7-dioxanonane derivatives: Study of crystallization mode and conformational dynamics in solution

Article abstract of DOI:10.1016/j.molstruc.2010.12.028

A homologous series of chiral bis-ketals and mixed hemiketal-ketals with rigid bicyclo[3.3.1]-3,7-dioxanonane skeleton has been prepared. The crystallization mode (conglomerate or racemic compound) was established by chiral HPLC with double UV/polarimetri

Full text of DOI:10.1016/j.molstruc.2010.12.028

Journal of Molecular Structure 989 (2011) 20–30
Contents lists available at ScienceDirect
Journal of Molecular Structure
journal homepage: www.elsevier.com/locate/molstruc
Chiral bicyclo[3.3.1]-3,7-dioxanonane derivatives: Study of crystallization mode
and conformational dynamics in solution
a,
⇑
a
a,c
a
a
Cornelia Uncu ßt a , Emeric Bartha , Drago sß Gherase , Ioan Andrei Loas , Florina Teodorescu ,
b
c
c
d
Richard A. Varga , Nicolas Vanthuyne , Christian Roussel , Ulf Berg
a
‘
‘
‘C.D. Nenitzescu’’ Center of Organic Chemistry, Spl. Independen ßt ei 202 B, 060023 Bucharest, Romania
‘Babe ßs -Bolyai’’ University, Faculty of Chemistry and Chemical Engineering, Cluj-Napoca, Romania
Université Paul Cézanne, UMR ISM2: Chirosciences, Marseille, France
Organic Chemistry, Lund University, PO Box 124, SE-22100 Lund, Sweden
b
c
d
a r t i c l e i n f o
a b s t r a c t
Article history:
A homologous series of chiral bis-ketals and mixed hemiketal-ketals with rigid bicyclo[3.3.1]-3,7-diox-
anonane skeleton has been prepared. The crystallization mode (conglomerate or racemic compound)
was established by chiral HPLC with double UV/polarimetric detection. The study of the solid-state struc-
Received 27 September 2010
Received in revised form 10 December 2010
Accepted 10 December 2010
Available online 15 January 2011
ture by single crystal X-ray diffraction disclosed interesting CHꢀ ꢀ ꢀ
p intermolecular interactions. Hinder-
1
ing of rotation of phenyl groups at (hemi)ketalic centres was found to occur. A H–DNMR study gave
#
ꢁ1
ꢁ1
activation barriers
DG of 15.9 kcal mol in the ketalic moiety and 10.3 kcal mol in the hemiketalic
Keywords:
X-ray analyses
Conglomerate
moiety. The experimental results were substantiated by MM and DFT calculations for ground and tran-
sition states.
Ó 2011 Published by Elsevier B.V.
CH/p interactions
DNMR
Hindered rotation barriers
DFT calculations
1
. Introduction
Spontaneous resolution of enantiomers upon crystallization
invoked for explaining symmetry breaking on crystallization as
well as spontaneous resolution of smaller (nanoscale) aggregates
[4,5].
The interplay between size, shape and conformational proper-
continues to be a challenging research area, under both practical
and theoretical stimuli. Separation of enantiomers is of greatest
interest to pharmaceutical industry, since more than 50% of phar-
maceutical active ingredients (including 9 of the 10 top drugs) and
almost 70% of drug candidates are chiral compounds [1,2]. Prefer-
ential crystallization of conglomerates [3] became a competitive
alternative to classical methods for enantiomeric resolution based
on diastereoselective interactions. Several other domains, such as
chiral catalysts, molecular materials with magnetic or electro-optic
properties are also exploiting crystalline conglomerates. A few rep-
resentative examples and applications may be found in a recent
critical review on spontaneous resolution [4].
At the present level of knowledge, the phenomenon of sponta-
neous resolution is still hard to explain and even harder to predict.
However, it seems almost certain now that the chemical tools
operating in supramolecular chemistry play a key-role in sponta-
neous resolution as well. Non-covalent bonding interactions be-
tween and within molecules are frequently and rationally
ties of molecules and the ‘‘supramolecular control’’ in spontaneous
resolution is better understood when dealing with series of com-
pounds which present a tendency towards symmetry breaking
on crystallization instead of isolates cases of conglomerates. Sev-
eral recent examples are families of chiral aryl glycerol ethers
and esters [6–8], 4-arenesulfonyliminocyclo-2-en-1-ones [9], gly-
colurils [10], with a much higher conglomerates frequency than
the poor occurrence of only 10% conglomerates in the bulk crystal
database. These families have been generated by operating minor
and regular variations, remotely from the stereogenic centres, on
the structural motif of a compound known to give spontaneous
resolution.
The obvious benefit of these studies in an enrichment of the
‘‘conglomerates pool’’ available for enantiomeric resolution by
preferential crystallization. Identification of specific intermolecular
interactions leading to supramolecular networks in crystal or of
the factors which might tip the balance between conglomerates
and racemic compounds are also valuable pieces of information is-
sued from these studies.
⇑
Corresponding author.
E-mail address: cuncuta@cco.ro (C. Uncu ßt a).
The present paper follows such an empirical approach, aiming
to design a series of chiral bicyclo[3.3.1]-3,7-dioxanonanes I with
0
022-2860/$ - see front matter Ó 2011 Published by Elsevier B.V.
doi:10.1016/j.molstruc.2010.12.028

C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
21
We succeeded in developing a ‘‘one pot’’ variant of the synthesis
at 0.1–0.5 mol scale with a reproducible yield in (±) 2 up to 65–
7
0%. Displacement of the diastereomeric equilibrium was due to
steady precipitation of (±) 2, both steps being reversible.
The mixed hemiketal–ketals (±) 3a–c were obtained by treating
(
±) 2 with methanol, ethanol or isopropanol at room temperature,
in chloroform solution and with catalytic amounts of sulfuric acid.
The reaction course was monitored by analytical tlc and has been
stopped when concentration of 3 began to decrease due to conver-
sion into bis-ketals 4. Pure derivatives 3a–c were isolated in mod-
est yields (10–15%) by preparative column chromatography on
silicagel. Derivative 3c presented a tendency towards dispropor-
tionation into 2 and 4c on standing in solution for several hours.
Reaction of (±) 2 with high excess of methanol or ethanol at re-
flux and with catalytic amounts of sulfuric acid provided bis-ketals
(
R,R'=H,Alkyl)
(R,R'=H,Me)
Scheme 1.
high conglomerate frequency. The starting structural motif was the
bis-diethylketal derivative I (R, R = Et), previously reported to give
a conglomerate on recrystallization from benzene [11]. Reasoning
that the propensity towards spontaneous resolution might origi-
nate in the molecular shape imprinted by the bicyclic core, we pre-
pared a series of C -symmetrical bis-ketals and C -hemiketal–
2 1
ketals congeners and studied their crystallization mode, looking
for new conglomerates. Using simple alkanols for ketalization,
0
(
±) 4a, (±) 4b in 83% and 91% yields, respectively. Analytically pure
samples were obtained by recrystallization from methanol (4a)
and ethanol (4b).
The reaction of (±) 2 with isopropanol performed as above gave
in 88% yield the triether 5 and only small amounts of 3c and 4c (2–
0
we operated steady size-variation of the alkyl groups R, R at the
4
% isolation yield).
periphery of the molecule, away from bridge-head chiral centres.
The crystallization mode was studied by single crystal X-ray dif-
fraction. Conglomerate spotting has been unambiguously achieved
by our previously reported methodology based on chiral HPLC with
double UV/polarimetric detection [12].
The tricyclic ether (±) 5 was obtained in 87% yield by refluxing a
solution of (±) 2 in benzene with catalytic amounts of p-toluene-
sulfonic acid.
Racemates 2, 3a–c, 4a–c and 5 were subjected to search proce-
dure for conglomerate detection.
0
From optically active bis-ether I (R = R = H), Chau and Corey
prepared optically active bicyclo[3.3.1]-3,7-diazanonanes(bispi-
dines) II [11], compounds which found lately interesting applica-
tions in asymmetric transformations [13]. The synthetic pathway
2.2. Conglomerate spotting by chiral HPLC double-monitored by UV/
CD detection
starts with the enantiomeric resolution of the conglomerate I
0
(
R = R = Et) by preferential crystallization. Successful results in
Prior analytical chiral HPLC experiments were performed in or-
der to find combination of eluting conditions and supports able to
discriminate the enantiomers of 2 and its derivatives 3, 4 and 5. All
the compounds except 4c were nicely separated, the best results of
this screening (retention times, retention factors k and enantiose-
our work, namely finding other conglomerates among the newly
prepared derivatives I, would enlarge the options in the first step
of this synthesis (see Schemes 1 and 2).
On recording the H and 13C NMR spectra of compounds I, an
exchange process occurring in the aromatic region was observed.
A DNMR study coupled with MM and DFT calculations disclosed
the origin and allowed determination of the activation energy of
the process.
1
lectivity factor a) being given in the Experimental part.
It is worth recalling that the methodology to search for con-
glomerate does not require the use of a chiral column yielding a
separation of the enantiomers. An achiral column can be used with
a dual detection, polarimeter/UV, polarimeter/IR, CD/UV when one
knows a priori that the enantiomers give a measurable polarimetric
or CD response in the mobile phase. The use of a chiral column
yielding complete separation of the enantiomers is safer since it
associates a separation technique with chirality detection. The
use of chiral support associated with polarimetric detection would
have been particularly rewarding in the case of metastable abnor-
mal conglomerates in which the cell contains an unequal mixture
of the enantiomers like those recently described by Bredikin et al.
[7].
Each ketalic or hemiketalic derivative was crystallized from the
alcohol used in its preparation and from benzene. Only four
compounds gave crystals suitable for the triage procedure: 3a
and 4a from methanol, 4b from benzene and 5 from isopropanol.
Derivatives 2, 3b, 3c and 4c gave crystals unsuitable for neither
conglomerate search procedure nor X-ray diffraction.
2
. Results and discussion
2
(
.1. Synthesis of bis-hemiketal 2, hemiketal–ketals 3 and bis-ketals 4
racemates)
The chiral bicyclo[3.3.1]-3,7-dioxanonane derivatives with gen-
eral formula I have been obtained as racemates in a remarkably
simple way from accessible, cheap materials. Firstly described by
Tilichenko [14] and afterwards by Corey and Chau [11], the reac-
tion of propiophenone with 37% aq. formaldehyde in ethanolic so-
dium hydroxide at 70 °C gave 1,5-diketone 1 (1:1 mixture of meso
and racemic diastereoisomers) in 95% yield. Further reaction of 1
with formaldehyde at 50 °C gave the bicyclic bis-hemiketal 2 (race-
mic) in 55% yield.
Scheme 2.

2
2
C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
inating the enantiomers. The results obtained are presented in
Fig. 1. Derivatives 3a, 4a and 5 were found to be racemic com-
pounds, both enantiomers being present in equal amounts in each
crystal. Conversely, each crystal of 4b was either (+) or (ꢁ) under
polarimetric detection, indicating a conglomerate [12].
5
. 8 6
5
. 2 5
2.3. Crystallization mode of derivatives 3, 4 and 5 studied by X-ray
diffraction
Derivatives 3a, 4a, 4b and 5 were analyzed by single crystal X-
ray diffraction. The crystallographic data in Table 1 show that 4b
crystallizes in the chiral space group P2 , confirming its con-
1 1 1
2 2
glomerate nature, whereas all the others belong to centro-symmet-
ric space groups, being therefore racemic compounds.
Fig. 2 displays the ORTEP representation of the molecules and
numbering of atoms; the dihedral angles significant for the confor-
mation of the rings and the orientation of the substituents are gi-
ven in Table 2.
3
. 6 1
The data show that the inner-core structure of the bis-ketals 4a,
4
b and of the hemiketal–ketal 3a are practically identical, the dif-
ferences between the significant dihedral angles being under 4°.
The heterocycle appears in chair conformation, with phenyl sub-
stituents in equatorial position and alkoxy/hydroxy groups in axial
position. We may conclude from this comparison that the con-
glomerate formation capacity of 4b is not dictated by the inner-
core shape; it remained unaltered under the peripheral substitu-
tion operated, but 3a and 4a congeners crystallized as racemic
compounds.
3
. 1 9
In the network of 3a, dimers consisting in pairs of enantiomers
are present. This association is based on two types of intermolecu-
lar interactions: (i) hydrogen bonding between hydroxyl group and
an endocyclic oxygen atom, with the following parameters:
0
0
0
d[(H(3 )ꢀ ꢀ ꢀO(1)] 2.078(4) Å, d [O(1)ꢀ ꢀ ꢀ(O(3 )] 2.864(7) Å, <[O(3 )–
0
H(3 )ꢀ ꢀ ꢀO(1)] 160.1(3)°; (ii) CHꢀ ꢀ ꢀ
p hydrogen bond between an
endocyclic CH group next to oxygen and an aromatic ring. The fol-
0
lowing parameters were measured: d[H(3B)ꢀ ꢀ ꢀC
g
(1 ) 3.0090(7) Å,
3
. 9 9
0
contact angle C(3)–H(3B) ? C
g
(1 ) 134.4(4)° (Fig. 3a). No further
3
. 4 6
interactions between dimers were found in the unit cell (Fig. 3b).
The importance of the CHꢀ ꢀ ꢀ
p
hydrogen bonding in controlling
the crystal packing in organic and organometallic compounds is
well documented by Nishio and co-workers in several reviews
[
15–17]. Systematic analyses of the Cambridge crystallographic
database were performed, using several parameters for surveying
the CH/ contacts. One of the statistically significant parameters
p
is the distance from the H atom to the closest carbon atom in the
aromatic ring (Datm), with an average value of 2.91 ± 0.12 Å for
intermolecular CH/
H(3B ) and C(11) measured in the crystal of 3a is 2.926 Å, a value
p contacts [18]. The distance Datm between
0
within the statistical range.
In the crystal of 4a, no close contacts between molecules have
been found. Apparently, only the thermodynamic factors are criti-
cal in determining the packing of molecules in crystal.
Chain-association based on intermolecular CHꢀ ꢀ ꢀ
p contacts
1
2 . 4 2
is formed in crystal of 4b, involving the CH group next to oxy-
0
gen. The following parameters were found: d[H(6A)ꢀ ꢀ ꢀC
g
(1 )
0
3
.0685(3) Å, contact angle C(6)–H(6A) ? C
g
(1 ) 134.2(2)°. The
0
0
distance Datm between H(6A ) and C(19) of 2.962 Å is within the
4
. 9 5
statistical range for CHꢀ ꢀ ꢀ
p
hydrogen bonding (Fig. 4a). The chains
are homochiral and spread along the b axis in crystal, without fur-
ther interchain interactions (Fig. 4b).
A comparison between the homochiral chains in conglomerate
4
b and the previously reported X-ray structure [11] of the enantio-
Fig. 1. Chiral HPLC analysis with double UV/polarimetric detection of individual
crystals in compounds 3a, 4a, 4b and 5.
0
pure (+)bis-2-bromoethylketal analog (I, R = R = CH
2
CH
2
Br) seems
appropriate. In the latter homomorphic crystal, chain-association
through CHꢀ ꢀ ꢀ interactions with quite similar parameters as
in 4b, namely d[H(2B)ꢀ ꢀ ꢀCg] 2.8954(3) Å, contact angle
A few crystals for each compound 3a, 4a, 4b, 5 were individu-
ally dissolved in ethanol and injected on the chiral column discrim-
p
=

C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
23
Table 1
Crystallographic data for compounds 3a, 4a, 4b and 5.
3
a
4a
4b
5
Empirical formula
Formula mass
Crystal system
Space group
a (Å)
C
22
H
24
O
4
C
23
H
28
O
4
C
25
H
32
O
4
C
21 22 3
H O
354.43
Monoclinic
P2 /n
13.524(4)
9.535(3)
15.964(5)
90
113.320(6)
90
368.45
Monoclinic
C2/c
18.455(3)
15.561(3)
15.294(3)
90
114.341(3)
90
396.51
Orthorhombic
P2
8.2763(9)
10.9626(12)
24.066(3)
90
90
90
322.39
Monoclinic
P2 /c
11.987(4)
12.069(4)
12.445(4)
90
110.200(5)
90
1689.6(9)
4
1
1
2
1
2
1
1
b (Å)
c (Å)
a
(°)
b (Å)
(°)
c
3
V (Å )
1890.4(11)
4
1.245
4001.5(12)
8
1.223
2183.5(4)
4
1.206
Z
D
calcd. (g/cm3)
1.267
F (0 0 0)
760
1584
856
688
Crystal size (mm)
0.39 ꢂ 0.30 ꢂ 0.23
0.23 ꢂ 0.16 ꢂ 0.09
0.34 ꢂ 0.32 ꢂ 0.19
0.30 ꢂ 0.23 ꢂ 0.14
) (mm 1)
ꢁ
0.084
0.082
0.080
0.084
l
(Mo K
a
h range (°)
1.68–25.00
17,746
3330 (Rint = 0.0929)
239
1.78–25.00
14,364
3528 (Rint = 0.0775)
248
2.00–25.00
15,871
2215 (Rint = 0.0557)
266
2.43–25.00
11,895
2972 (Rint = 0.0476)
219
No. of reflections collected
No. of independent reflections
No. of parameters
Absorption correction
Multi-scan (Bruker SAINT)
0.1401
R
1
[I > 2r
(I)]
0.1011
0.0523
0.0679
wR
2
0.2251
1.295
0.236/ꢁ0.196
0.1680
1.193
0.179/ꢁ0.189
0.1097
1.222
0.155/ꢁ0.170
0.1296
1.149
0.170/ꢁ0.251
GOF on F2
3
Largest difference electron density (e/Å )
3
a
4a
4
b
5
Fig. 2. ORTEP diagram of compounds 3a, 4a, 4b and 5.
C(2)–H(2B) ? C
g
133.36°, may be observed. Additional close con-
group (Fig. 6a). The CHꢀ ꢀ ꢀ
p interactions involving the former link
molecules of the same configuration, whereas the latter links
0
tacts through CHꢀ ꢀ ꢀBr interactions with d[H(13A)ꢀ ꢀ ꢀBr(1 )] =
2
.9493(3) Å are present (Fig. 5).
molecules of opposite configurations. The following parameters
0
000
In the crystal structure of the racemic compounds 5, each mol-
of the CHꢀ ꢀ ꢀ
p
interactions were found: d[H(6B)ꢀ ꢀ ꢀC
g
(2 )]
0
000
ecule interacts with four neighboring molecules through CHꢀ ꢀ ꢀ
p
2.9482(9) Å, contact angle C(6)–H(6B) ? C
g
(2 ) 131.2(2)° and
0
00
000
interactions involving two different endocyclic CH groups: one is
d[H(7B)ꢀ ꢀ ꢀC
g
(1 )] 2.8296(7) Å, contact angle C(7)–H(7B) ? C
g
(1 )
0
that next to oxygen (as before) and the other is the bridge-CH
129.0(1)°. The Datm values are 3.111 Å [distance between H(7 B)

2
4
C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
Table 2
The X-ray structure of enantiopure (ꢁ) 5 has been previously
published by Corey and Chau [11]. Fig. 7 displays the elemental cell
of the enantiomorphous crystal (space group P2 ). We did not de-
Dihedral angles in compounds 3a, 4a, 4b and 5.
Dihedral angle
3a
4a
4b
5
1
tect any intermolecular interactions in the crystal.
O
O
O
O
1
1
2
2
C
C
C
C
O
O
O
O
C
C
C
C
2
3
5
6
1
1
2
2
C
C
C
C
C
C
C
C
1
4
4
1
2
2
5
5
C
C
C
C
C
O
3
C
O
7
ꢁ51.5
ꢁ52.4
ꢁ54.4
ꢁ36.3
ꢁ44.9
ꢁ37.6
ꢁ45.5
7
58.7
ꢁ51.3
58.0
57.4
ꢁ53.6
57.9
56.5
ꢁ54.0
55.4
Summing-up, the outcome in the crystallization mode of the
chiral bicyclo[3.3.1]-3,7-dioxanonanes was conglomerate (4b) or
racemic compound (3a, 4a, 5). In 4b, formation of homochiral
7
7
C
C
C
C
C
C
C
C
3
10
172.0
ꢁ67.9
172.3
ꢁ65.9
146.8
27.3
172.7
ꢁ66.9
174.4
ꢁ64.0
ꢁ26.9
ꢁ147.0
ꢁ26.5
ꢁ146.8
175.0
ꢁ64.4
174.3
ꢁ66.1
ꢁ25.4
ꢁ145.8
ꢁ29.3
ꢁ148.8
ꢁ160.4
ꢁ43.4
159.9
–
chains occurs, based on CHꢀ ꢀ ꢀ
p intermolecular hydrogen bonding.
3
6
6
However, this interaction does not seem to be a driving force to-
wards symmetry breaking on crystallization. Although 4a is the
closest analog of 4b, it crystallized as racemic compound and did
not present any intermolecular interactions in the network.
Formation of intimate pairs of enantiomers in the racemic
crystal of 3a may be explained by double interactions through
16
4
11
11
17
17
10
10
16
16
C
2
2
5
5
O
1
3
2
4
151.4
34.9
148.3
–
C
C
C
O
O
O
ꢁ33.5
ꢁ154.6
OHꢀ ꢀ ꢀO and CHꢀ ꢀ ꢀ
p
hydrogen bonding. Double interactions through
contacts were also found in the tricyclic
0
0
and C(11)] and 3.071 Å [distance between H(B ) and C(20)], both
two different CHꢀ ꢀ ꢀ
p
values being out of the statistical range for CHꢀ ꢀ ꢀ
p
hydrogen bond-
derivative 5; it leads to formation of homochiral layers, but race-
mic crystals result since adjacent layers have opposite handedness.
ing in the crystals.
This arrangement leads to a two dimensional double-layer; the
layers are homochiral, with alternating handedness and are
stacked along the a axis (Fig. 6b).
2
A few carbocyclic analogs of derivatives I, namely C -symmetric
endo, endo- bicyclo[3.3.1]nonane-2,6-diols gave a similar variety in
crystallization mode. Different hydrogen bonded supramolecular
Fig. 3. (a) View of dimer association based on intermolecular Oꢀ ꢀ ꢀH bonds (only hydrogens involved in intermolecular interactions are shown) and C–Hꢀ ꢀ ꢀ
p contacts in the
0
crystal of 3a [symmetry equivalent atoms (1 ꢁ x, ꢁy, ꢁz) are given by ; Cg(1) is the centre of the benzene ring C10–C15]. (b) View of the unit cell of 3a.
Fig. 4. (a) View of chain-association based on intermolecular C–Hꢀ ꢀ ꢀ
symmetry equivalent atoms (1 ꢁ x, ꢁ0.5 + y, 1.5 ꢁ z) and (1 ꢁ x, 0.5 + y, 1.5 ꢁ z) are given by and , respectively; Cg(1) is the centre of the benzene ring C16–C21]. (b) Crystal
packing of 4b along the a axis.
p contacts (only hydrogens involved in intermolecular interactions are shown) in the crystal of 4b
0
00
[

C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
25
synthons were identified, leading to crystallization as conglomer-
ate, as racemic compounds with alternating (+) and (ꢁ) enantio-
mers or racemic crystals containing enantiomers separated into
discrete domains [19].
1
2
.4. Conformational dynamics by H–DNMR and computational study
1
2
.4.1. Dynamic H NMR spectroscopy
The room temperature 1H and 13C NMR spectra of hemiketal–
ketals 3a–c and bis-ketals 4a–c display selective broadening of
the signals in the aromatic region, indicating exchange processes.
This feature was not found in the spectra of the parent bis-hemike-
tal 2. A closer look revealed that compounds 3 experienced broad-
ening for only
a part of the aromatic signals, whereas in
compounds 4 all aromatic signals were affected. The other signals
in the spectra lacked any broadening.
Fig. 7. Elemental cell in the enantiomorphous crystal of (ꢁ) 5 [11].
This behaviour called for a more detailed investigation. For this
1
purpose, we examined the H NMR spectra (500 MHz) of the
homologous compounds 2, 3a and 4a at variable temperature, in
tetrahydrofurane-d
8
as solvent. In the ꢁ70 to 50 °C temperature
range, the bis-ketal 4a underwent broadening of the signals in
the domain 7.2–8.1 ppm. The bis-hemiketal 2 behaved similarly,
but in a much lower temperature range (ꢁ70 ° to ꢁ10 °C). The
spectrum of the mixed compound 3a appeared as superposition
of the spectra of 2 and 4a. Selected variable-temperature spectra
are presented in Fig. 8.
Fig. 5. View of chain-association through intermolecular CHꢀ ꢀ ꢀ
p
contacts and
0
CHꢀ ꢀ ꢀBr interactions in bis-2-bromoethylketal (I, R = R = CH
2 2
CH Br) [11].
Fig. 6. (a) View of C–Hꢀ ꢀ ꢀ
.5 ꢁ y, ꢁ0.5 + z), (1 ꢁ x, ꢁ0.5 + y, 0.5 ꢁ z), (x, 0.5 ꢁ y, 0.5 + z) and (1 ꢁ x, 0.5 + y, 0.5-z) are given by ,
Cg(2) is the centre of the benzene ring C16–C21]. (b) Crystal packing in racemic compound 5 along the c axis.
p intermolecular interactions (only hydrogens involved in intermolecular interactions are shown) in the crystal of 5 [symmetry equivalent atoms (x,
0
00 000 0000
and respectively; Cg(1) is the centre of the benzene ring C10–C15,
0
,

2
6
C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
Fig. 8. Variable temperature 1H NMR (500 MHz) spectra in 2
00 K temperature range (the inset at 225 K is the experimental spectrum after adding trifluoroacetic acid). For the simulated spectra in the exchange domain, the rate
constant values are given. Right column shows experimental spectra of 3a in the 200–325 K temperature range.
8
H -THF. Left column shows experimental (bottom) and calculated (upper) spectra of compound 2 in the 200–
3

C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
27
Table 3
In order to find experimental support for this idea, a droplet of
wet trifluoroacetic acid was added to the sample of 2, and then the
NMR spectrum at the coalescence temperature (225 K) was
recorded. No change in the spectrum was observed (see inset in
Fig. 8). Should the ring-opening/ring-closure mechanism be
involved, a catalytic effect on the reaction rate would be expected
and, accordingly, a change of the NMR pattern should occur in the
above experiment. We concluded therefore that the exchange
process observed in the DNMR experiment is due to steric
hindrance of the phenyl rotation occurring with different rates at
ketalic and hemiketalic centers.
#
Rotational barrier of the phenyl group at the ketalic center (DG1 ) and the hemiketalic
center (DG2 ).
#
#
#
#
#
#
Comp. Temperature
K)
D
G1
D
G2
DD
G
¼
D
G1
ꢁ
DG2
(
ꢁ1
ꢁ1
ꢁ1
(
kcal mol
)
(kcal mol
)
(kcal mol )
2
3
225
225
3
–
–
10.3
10.4
–
a
a
25
15.8
15.9
5.4
5.6
4
325
–
In order to better understand the unusually large difference
#
#
between
D
G1 and
D
G2 values,
a
computational study was
undertaken.
2
.4.2. Computational study of conformation in ground and transition
states
The calculations were performed for the C
2
-symmetrical mole-
cules 2 and 4a; comparison between results for 2 and 4a would ac-
count for the effects of replacing OH by OMe on ground and
transition states along torsion of phenyl group as reaction
coordinate.
In order to identify low-energy conformations significant for the
ground state of the molecule, a stochastic conformational search
was carried out at molecular mechanics level. The MM calculations
⁄
⁄
were performed using the MM3 , AMBER and OPLS force fields of
the Macromodel program version 6.5 [24], using default parame-
ters. All MM force fields gave congruent results.
It turned out that only one conformation was competitive in the
ground state of 2 as well as of 4a. The next low-energy conformer
ꢁ1
is above the lowest-energy conformer by 7.20 kcal mol in 2 and
ꢁ
1
7
.75 kcal mol in 4a.
Fig. 9. Energy profile for phenyl rotation in 4a calculated by the MM2 force field in
the MacMimic program.
Next, a dihedral drive computation was performed in order to
examine the rotational path and to find an approximate transition
state. The MM2(91) force field implemented in the MacMimic pro-
gram package [25] was used. This force field was chosen as it has a
more advanced way to accommodate the torsions of the phenyl
ring in the strained transition state. The rotational energy profile
against the dihedral angle C17–C16–C5–O2 (numbering from
Fig. 2) for compound 4a is shown in Fig. 9.
Finally, the transition state geometries for 2 and 4a were local-
ized by DFT calculations, starting from the ‘‘transition state’’ geom-
etries obtained in the MM results. Density functional theory
computations were performed with the ADF program package [26].
A proper transition state was straightforwardly established for
compound 2, but compound 4a raised considerable problems as
multiple negative eigen values were obtained, corresponding to
methyl group rotations, etc. Table 4 displays the results of DFT cal-
culations for ground and transition states in compounds 2 and 4a.
Not unexpectedly, the results differ between various functionals.
Comparison between calculated and experimental results for
the ground state was rewarding: an overlay of the calculated and
The experimental results may be explained by hindered rota-
tion of the equatorial phenyl groups past the axial methoxy or hy-
droxyl substituents. The NMR pattern was analyzed as an exchange
0
0
process of the type ABCDE ? AA BB E, expected for rotation of the
phenyl group between two identical conformations. The dynamic
NMR experiments were performed as described by Sandström
[
20]. The rate constants were evaluated by visual fitting of the
experimental spectra to spectra calculated by the program g-
NMR [21]. A very good fit between experimental and simulated
spectra was obtained (see Fig. 8, left column), allowing determina-
tion of the rate constant (k) values. Calculation of the free energy of
#
ꢁ1
activation gave
D
G1 = 15.9 kcal mol
(at 325 K) for the phenyl
rotation occurring at the ketalic center (in 4a and in the ketalic
#
ꢁ1
moiety of 3a) and DG2 = 10.3 kcal mol (at 225 K) for the rotation
occurring at the hemiketalic center (in 2 and in the hemiketalic
moiety of 3a, Table 3).
This difference in sterical rotational barriers of 5.5 ±
ꢁ1
0
.1 kcal mol
is quite large, considering that Me/H change on
crystal geometry of 4a gave almost perfect superposition (Fig. 10).
#
E
oxygen occurs a couple of bonds away and, further, that both
methoxy/hydroxy groups can rotate for avoiding severe steric
As far as calculated activation energies
D
are concerned, a di-
rect comparison with experimental data is not appropriate, since
#
interactions on phenyl rotation.
only
DG
values could be accurately determined in the DNMR
#
DG
An alternative explanation for this large difference in
experiment (the appropriate comparison should be with
#
values might reside in a different mechanism operating at the
hemiketalic center. ring-opening/ring-closure reaction may
experimental activation enthalpy
D
H ). However, the differences
#
#
A
between calculated activation energies
DE1 and
DE2 and experi-
#
#
occur, passing through an open-chain intermediate (5-hydroxy
ketone). Provided a free phenyl rotation in this intermediate,
coalescence of the paired signals in the aromatic region would be
expected. A ring-opening, ring-closure process is well documented
in saccharides in which mutarotation is occurring in neutral
medium or can be accelerated by acidic, basic or bifunctional cat-
alysts [22,23].
mental barriers
D
G1 and
D
G2 should be comparable. Indeed, the
barriers being of steric origin at both hemiketalic and ketalic
centres (as undoubtedly established), the entropic contribution in
#
calculation of DD
G
is expected to cancel. The data in Tables 3
#
ꢁ1
and 4 show that the experimental DD
agreement with the DD
G
= 5.5 kcal mol is in good
#
ꢁ1
E
value of 6.1 kcal mol calculated with
BLYP functional. The simpler functionals gave poorer results.

2
8
C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
Table 4
#
#
Ground state (GS) and transition state (TS) energies from DFT calculations; calculated activation energy on phenyl group rotation at the ketalic (DE1 ) and hemiketalic (DE2
)
centres.
#
ꢁ1
#
ꢁ1
#
#
#
ꢁ1
Comp.
Method
Energy (au)
D
E1 (kcal mol
)
D
E2 (kcal mol
)
DDE
¼
D
E1
ꢁ
DE2 (kcal mol )
4
4
2
2
a TS
a GS
TS
LDA/sz
LDA/dz
BLYP + LDA + GGA
ꢁ15.45970141
ꢁ12.75952578
ꢁ13.73201992
ꢁ15.47928809
ꢁ12.78671203
ꢁ13.75332995
ꢁ14.67161886
ꢁ12.15132781
ꢁ13.03145735
ꢁ14.68800121
ꢁ12.17457825
ꢁ13.04295997
12.29
17.06
13.37
2.01
2.47
6.15
LDA/sz
LDA/dz
BLYP + LDA + GGA
LDA/sz
LDA/dz
BLYP + LDA + GGA
10.28
14.59
7.22
GS
LDA/sz
LDA/dz
BLYP + LDA + GGA
We observe that the BLYP + LDA + GGA functional satisfactorily
reproduced both ground state structures and rotational barriers,
which was actually the aim of this computational study. The rea-
son for this probably lies in the fact that these molecules have only
one possible and relatively rigid conformation according to both
experimental and computational results. In this conformation the
O–Me(H) groups are axial and thus a finer tuning of the anomeric
effect becomes less important. On the other hand, the LDA/sz and
LDA/dz level of theory failed to reproduce the rotational barrier.
The origin of the relatively large difference in barrier is best
visualized by following the rotation transit in MM computations.
Fig. 11 shows that in the bis-ketal 4a there is a transanular steric
clash between one ortho hydrogen and methylene hydrogen and
between the other ortho hydrogen and the methyl in methoxy
group. The latter interaction is missing in the transition state of 2.
In conclusion, the MM force fields and the higher order DFT cal-
culations of the steric barriers to rotation of the phenyl group in 2
and 4a gave a reasonable agreement with the DNMR experiments.
The barrier is much lower in the bis-hemiketal compound 2 than in
the bis-ketal compound 4a on steric ground. These results ruled
out the participation of a ring-chain equilibrium in 2 to account
for the observed large difference in the barriers. The activation en-
ergy of the ring-chain tautomerism in the bis-hemiketal 2 is prob-
ably too large to play any role in the exchange process. However, it
might be operating in the case of substitution in ortho-positions of
the phenyl groups.
3
. Conclusion
We undertook a study on the crystallization mode (conglomer-
ate vs racemic compounds) of several chiral bis-ketals and mixed
hemiketal–ketals with bicyclo[3.3.1]-3,7-dioxanonane skeleton.
Conglomerate spotting has been unambiguously achieved by chiral
HPLC with double UV/CD monitoring. The single crystal X-ray anal-
ysis enabled identification of intermolecular interactions in the
Fig. 10. Superposition of DFT calculated ground state and X-ray geometry of 4a.
network. A supramolecular synthon involving CHꢀ ꢀ ꢀ
p interactions
was found in the homochiral chains of conglomerate 4b.
Steric hindrance to rotation of equatorial phenyl substituents
was found to occur on the NMR time-scale. The rotational barriers
at both hemiketalic and ketalic centres were measured by 1H–
DNMR and were compared with theoretical barriers obtained by
MM and DFT calculations. A fairly good agreement between exper-
iment and calculations was obtained using BLYP + LDA + GGA
functional.
4
. Experimental part
4.1. Instrumentation and reagents
Melting points were determined on a Boetius hot plate and are
uncorrected. The elemental analyses were performed with an ECS
010 COSTECH instrument. The NMR spectra were recorded with
4
1
a Varian Gemini 300BB instrument (operating at 300 MHz for
and at 75 MHz for C), in CDCl
H
1
3
3
solution, at room temperature.
Fig. 11. Selected van der Waals spheres of transition geometry of 4a.
The chemical shifts (d) values are given in ppm from internal

C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
29
TMS, the coupling constants J are in Hz. The abbreviation brd sgn
J = 7.0), 7.81 (4H, d, J = 6.9); 13C NMR: 19.8, 36.8, 37.1, 67.3, 99.9,
127, 127.5, 127.8, 142. Chiral HPLC analysis: Chiralcel OD-H, Hex-
ane/2-PrOH (90/10), 1 mL/min, 25 °C, UV detection at 254 nm and
polarimeter, Rt(ꢁ) = 3.95, Rt(+) = 5.39, k(ꢁ) = 0.30, k(+) = 0.77,
1
stands for broad signals. In the dynamic NMR experiments,
NMR spectra were recorded with a Brucker AMX 500 MHz spec-
trometer. Typical solution was 8 mM in 0.7 mL THF-d , and the sol-
H
8
vent peak was used as internal shift standard. Analytical TLC was
performed with silica gel F254 (Merck) on aluminum strips
a = 2.58.
(
10 cm length). For column chromatography, silica gel 60 (Merck)
4.4. Preparation of hemiketal–ketals 3 and of bis-ketals 4
was used. The elution solvents are abbreviated as follows: diethyl
ether (EE), petroleum ether/boiling range 40–67 °C (PE).
The analytical chiral HPLC analyses were performed on an unit
composed of Merck D-7000 system manager, Merck-Lachrom L-
To 0.68 g (2 mmol) 2 in 2.5 mL chloroform was added 0.16 mL
(4 mmol) methanol, then 0.16 mL of 0.25 M methanolic solution
of sulfuric acid (0.04 mmol sulfuric acid) and the mixture was stir-
red at room temperature for 5 min. On diluting with 5 mL methy-
lene chloride, the unreacted 2 partly precipitated and was filtered
off (0.124 g). The filtrate was handled by column chromatography,
giving in the elution order: 4a (0.163 g, 22% yield) eluted with 4%
EE/PE, 3a (0.121 g, 17% yield) eluted with 10% EE/PE and unreacted
2 (0.280 g) eluted with 50% EE/PE.
7
100 pump, Merck-Lachrom L-7200 autosampler, Merck-Lachrom
L-7360 oven, Merck-Lachrom L-7400 UV-detector and Jasco OR-
590 polarimeter, with Chiralcel OD-H (250 ꢂ 4.6 mm) or Chir-
alpak AD (250 ꢂ 4.6 mm). Retention times Rt in minutes, retention
factors k = (Rt )/Rt and enantioselectivity factor = k /k are
ꢁ Rt
given. Rt was determined by injection of tri-t-butyl benzene.
1
i
i
0
0
a
2
1
0
By performing the reaction at room temperature with ethanol
instead of methanol and with similar work-up, compounds 3b
and 4b were obtained in 10% and 28% yield, respectively.
The reaction with isopropanol was performed with high excess
of alcohol (1: 20 M ratio), with stirring at room temperature for
72 h. Working-up as above, 3c and 4c were obtained in 12% and
5.5% yield, respectively. It was noted that 3c had a tendency to-
wards disproportionation into 2 and 4c on standing in solution
for several hours.
Working with high excess of methanol or ethanol (1: 20 M ra-
tio) and heating at reflux for 15–30 min, the bis-ketals 4a and 4b
were isolated in 83% and 91% yields, respectively (at 0.05 mol
scale).
4
.2. X-ray data collection and structure determination
Data were collected on Bruker SMART APEX diffractometer
‘‘Babes-Bolyai’’ University, Cluj, Romania), using graphite-mono-
chromated Mo K radiation (k = 0.71073 Å). For this purpose the
(
a
crystals were mounted on cryoloops and the data were collected
at room temperature (297 K). The structure was solved by direct
methods SHELXS-97 and successive difference Fourier syntheses
2
and refined against F on all data by full-matrix least-squares with
SHELXL-97 [27]. All non-hydrogen atoms were anisotropically re-
fined. Hydrogen atoms were placed at idealized positions with iso-
tropic thermal parameters set at 1.5 times for the methyl
hydrogens and 1.2 for the rest, respectively, of the carbon atom
to which they were attached. The methyl groups were allowed to
rotate but not to tip. Atomic scattering factors for neutral atoms
and real and imaginary dispersion terms were taken from Interna-
tional Tables for X-ray Crystallography [28]. The drawing was cre-
ated with the Diamond program [29].
Crystallographic data (excluding structure factors) for the
structures in this paper have been deposited with the Cambridge
Crystallographic Data Centre as supplementary publication num-
bers CCDC 734925–734928. Copies of the data can be obtained,
free of charge, on application to CCDC, 12 Union Road, Cambridge
CB2 lEZ, UK [fax: +44(0) 1223 336033 or e-mail:deposit@
ccdc.cam.ac.uk].
(±) 6-Methoxy-1,5-dimethyl-2,6-diphenyl-3,7-dioxa-bicyclo-
[3.3.1]-nonane-2-ol 3a: m.p. 179–184 °C (methanol); Anal. Calcd.
for C22
3362 cm
H
26
O
;
4
: C, 74.55, H, 7.39. Found: C, 74.34, H, 7.57; IR:
ꢁ
1
1
H NMR: 0.72 (3H, s), 0.74 (3H, s), 2.02 (1H, dd,
4
4
J = 12.5, J = 2.2), 2.08 (1H, dd, J = 12.5, J = 2.2), 2.45 (1H, s, OH),
3.06 (3H, s), 3.19 (1H, dd, J = 11.7, J = 2.2), 3.26 (1H, d, J = 11.7),
3.38 (1H, dd, J = 11.7, J = 2.2), 3.42 (1H, d, J = 11.7), 7.30–8.00 (10
4
4
H, m, brd sgn). Chiral HPLC analysis: Chiralcel OD-H, Hexane/2-
PrOH (99/1), 1 mL/min, 25 °C, UV detection at 254 nm and polar-
imeter, Rt(ꢁ) = 4.95, Rt(+) = 12.42, k(ꢁ) = 0.65, k(+) = 3.14,
a = 4.83.
(±) 6-Ethoxy-1,5-dimethyl-2,6-diphenyl-3,7-dioxa-bicyclo-
[3.3.1]-nonane-2-ol 3b: m.p. 184–190 °C (ethanol); Anal. Calcd.
for C23
H
28
O
;
4
: C, 74.97, H, 7.66. Found: C, 74.80, H, 7.30; IR:
ꢁ
1
1
3
389 cm
H NMR: 0.74 (3H, s), 0.75 (3H, s), 1.19 (3H, t,
4
4.3. ‘‘One pot’’ synthesis of bis-hemiketal 2
J = 7.1), 2.02 (1H, dd, J = 12.5, J = 2.6), 2.13 (1H, dd, J = 12.5,
4
J = 2.6), 2.38 (1H, s, OH), 3.19 (1H, dq, J = 9.5, J = 7.1), 3.19 (1H,
4
In a 1L round-bottomed flask provided with reflux condenser,
dd, J = 11.7, J = 2.6), 3.27 (1H, d, J = 11.7), 3.31 (1H, dq, J = 9.5,
J = 7.1), 3.38 (1H, dd, J = 11.7, J = 2.6), 3.44 (1H, d, J = 11.7), 7.30–
4
dropping funnel and inner thermometer were introduced in order
5
6
1
3
g (0.125 mol) sodium hydroxide, 5 mL water, 240 mL ethanol and
7.09 g (66.4 mL, 0.5 mol) propiophenone, under magnetic stirring.
8.00 (10 H, m, brd sgn); C NMR: 15.1, 19.8, 19.9, 36.7, 37.4,
37.5, 56.4, 67.8, 68.2, 100.3, 103.4, 126.7 (brd sgn), 127.5, 127.6,
128.1, 129.5 (brd sgn), 138.1, 141.7. Chiral HPLC analysis: Chiralcel
OD-H, Hexane/2-PrOH (99/1), 1 mL/min, 25 °C, UV detection at
254 nm and polarimeter, Rt(ꢁ) = 4.18, Rt(+) = 6.94, k(ꢁ) = 0.37,
The mixture was heated up to 70 °C, then 37% aq. formaldehyde
18.8 mL, 0.25 mol) was added dropwise over 45 min. Stirring at
(
7
5
1
0 °C was continued for 60 min, then the mixture was cooled to
0 °C and a second portion of 37% aq. formaldehyde (75.2 mL,
mol) was added quickly. After 15 min, a white solid started to
k(+) = 1.28,
a = 3.43 and Rs = 1.48.
(±) 6-Isopropoxy-1,5-dimethyl-2,6-diphenyl-3,7-dioxa-bicyclo-
precipitate. Heating at 50 °C and stirring continued for 90 min,
then the mixture was left overnight. The precipitate was filtered
off, washed with 100 mL water and air-dried, providing a first crop
of 46.6 g bis-hemiketal 2. On leaving the filtrate overnight at room
temperature, a second crop of 13.5 g 2 was get, corresponding to
[3.3.1]-nonane-2-ol 3c: m.p. 168–172 °C (ethanol); Anal. Calcd.
30 4
for C24H O : C, 75.36, H, 7.91. Found: C, 75.62, H, 7.49; IR:
ꢁ
1
1
3398 cm
;
H NMR: 0.76 (3H, s), 0.82 (3H, s), 1.12 (3H, d,
4
J = 6.1), 1.15 (3H, d, J = 6.1); 2.01 (1H, dd, J = 12.4, J = 2.5), 2.12
(1H, dd, J = 12.4, J = 2.5), 2.20 (1H, s, OH), 3.03 (1H, dd, J = 11.5,
J = 2.5), 3.38 (1H, dd, J = 11.5, J = 2.5), 3.50 (2H, dd, J = 11.5,
J = 2.5), 3.75 (1H, hpt, J = 6.1), 7.30–7.55 (7 H, m), 7.79 (2H, d,
4
4
4
4
7
0.7% overall yield.
±) 2,6-Dihydroxy-1,5-dimethyl-2,6-diphenyl-3,7-dioxa-bicy-
(
clo[3.3.1]-nonane 2: m.p. 207 ꢁ 9 °C (lit. m.p. 184 ꢁ 5 °C [11]; it
J = 6.8), 8.06 (1H, d, J = 7.4). Chiral HPLC analysis: Chiralcel OD-H,
Hexane/2-PrOH (99/1), 1 mL/min, 25 °C, UV detection at 254 nm
and polarimeter, Rt(ꢁ) = 4.02, Rt(+) = 5.98, k(ꢁ) = 0.32, k(+) = 0.96,
should be noted that our sample presented only a change in crysta-
1
linity around 185 °C); H NMR: 0.73 (6H, s); 2.10 (2H, s), 3.32 (2H,
d, J = 11.6), 3.59 (2H, d, J = 11.6), 7.32 (2H, t, J = 6.9), 7.39 (4H, t,
a = 3.03 and Rs = 0.98.

3
0
C. Uncu ßt a et al. / Journal of Molecular Structure 989 (2011) 20–30
(
±) 2,6-Dimethoxy-1,5-dimethyl-2,6-diphenyl-3,7-dioxa-bicy-
clo[3.3.1]-nonane 4a: m.p. 140–141 °C (methanol); Anal. Calcd.
for C23
: C, 74.97, H, 7.66. Found: C, 74.74, H, 7.71; 1H NMR:
.75 (6H, s), 2.05 (2H, s), 3.07 (6H, s), 3.23 (2H, d, J = 11.6); 3.30
(Project Number PNII – IDEI 53/2007) is gratefully acknowledged
(C.U., E.B., D.G.). Thanks are due to the NATIONAL CENTER FOR
X-RAY DIFFRACTION (‘‘Babes-Bolyai’’ University, Cluj-Napoca,
Romania) for the support in the solid-state structure determina-
tions (R.A.V.).
28 4
H O
0
(
2H, d, J = 11.6), 7.30–7.38 (6H, brd sgn), 7.44–7.54 (2H, brd sgn),
7
1
1
.94–8.20 (2H, brd sgn); 13C NMR: 19.8, 37.3, 37.6, 49.0, 68.2,
03.6, 126.9 (brd sgn), 127.7 (brd sgn), 127.8, 128.3 (brd sgn),
29.8 (brd sgn), 137.4. Chiral HPLC analysis: Chiralpak AD, Hex-
Appendix A. Supplementary material
ane/2-PrOH (99/1), 1 mL/min, 25 °C, UV detection at 254 nm and
polarimeter, Rt(ꢁ) = 3.46, Rt(+) = 3.99, k(ꢁ) = 0.14, k(+) = 0.31,
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.molstruc.2010.12.028.
a
= 2.27 and Rs = 2.04.
(
±) 2,6-Diethoxy-1,5-dimethyl-2,6-diphenyl-3,7-dioxa-bicyclo-
References
[
[
2
3.3.1]-nonane 4b: m.p. 172–8 °C (ethanol) (lit. m.p. 176–8 °C)
11]. The enantiopure crystals of (+) 4b or (ꢁ) 4b (benzene): m.p.
[1] P. Piras, C. Roussel, J. Pharmaceut. Biomed. Anal. (2008) 839–847.
[2] Y. Wang, A.M. Chen, Org. Process Res. Dev. 12 (2008) 282–290.
1
16–7 °C (lit. m.p. 206–8 °C, sealed tube) [11]; H NMR: 0.76 (6H,
[
[
3] G. Coquerel, Top. Curr. Chem. 269 (2007) 1–51.
4] L. Perez-Garcia, D.B. Amabilino, Chem. Soc. Rev. 36 (2007) 942–967.
s), 1.19 (6H, t, J = 7.0), 2.09 (2H, s), 3.13–3.36 (8H, m), 7.30–8.00
(
1
1
1
3
10H, m, brd sgn); C NMR: 15.1, 19.9, 37.2, 37.7, 56.4, 68.3,
[5] L. Perez-Garcia, D.B. Amabilino, Chem. Soc. Rev. 31 (2002) 342–356.
[
6] A.A. Bredikhin, Z.A. Bredikhina, D.V. Zakharychev, L.V. Konoshenko,
Tetrahedron: Asymmetry 18 (2007) 1964–1970.
03.4, 126.8 (brd sgn), 127.5 (brd sgn), 127.7, 128.2 (brd sgn),
29.6 (brd sgn), 138.2. Chiral HPLC analysis: Chiralpak AD,
[
7] A.A. Bredikhin, Z.A. Bredikhina, V.G. Novikova, A.V. Pashagin, D.V. Zakharychev,
A.T. Gubaidullin, Chirality 20 (2008) 1092–1103.
Hexane/2-PrOH (99/1), 1 mL/min, 25 °C, UV detection at 254 nm
and polarimeter, Rt(ꢁ) = 3.19, Rt(+) = 3.61, k(ꢁ) = 0.05, k(+) = 0.19,
[
[
8] A.A. Bredikhin, Z.A. Bredikhina, Book of abstracts, 2008, p. 83.
9] R.G. Kostyanovsky, A.P. Avdeenko, S.A. Konovalova, G.K. Kadorkina, A.V.
Prosyanik, Mendeleev Commun. 10 (2000) 16–18.
a
= 3.72.
(
±) 2,6-Diisopropoxy-1,5-dimethyl-2,6-diphenyl-3,7-dioxa-bi-
[10] K.A. Lyssenko, D.G. Golovanov, A.N. Kravchenko, I.E. Chikunov, O.V. Lebedev,
N.N. Makhova, Mendeleev Commun. 14 (2004) 105–107.
[11] F.H. Chau, E.J. Corey, Tetrahed. Lett. 47 (2006) 2581–2583.
cyclo[3.3.1]-nonane 4c: m.p. 140–143 °C (isopropanol); Anal.
1
Calcd. for C27
NMR: 0.81 (6H, s), 1.09 (6H, d, J = 6.1), 1.10 (6H, d, J = 6.1), 2.02
2H, s), 3.04 (2H, d, J = 11.5), 3.39 (2H, d, J = 11.5), 3.71 (2H, hpt,
J = 6.1), 7.30–7.55 (8H, m, brd sgn), 8.06 (2H, d, J = 7.8); C NMR:
0.7, 24.2, 24.3, 37.7, 38.0, 65.7, 68.9, 104.5, 126.2, 127.5, 128.0,
29.5, 139.4. Chiral HPLC analysis: no enantiomeric separation
has been achieved on Chiralpak AD, Chiralpak AS, Chiralpak IC,
Chiralcel OD-H, Chiralcel OJ, Chiralcel OB-H, Ulmo (S,S) and
Whelk-O1 (S,S), with Hexane/2-PrOH (99/1).
36 4
H O : C, 76.38, H, 8.55. Found: C, 76.28, H, 8.42; H
[
[
12] C. Roussel, N. Vanthuyne, J.L. Jobert, A.I. Loas, A.E. Tanase, D. Gherase, Chirality
19 (2007) 497–502.
13] M. Breuning, Synthesis (2008) 2841–2867.
(
13
[14] M.N. Tilichenko, Zh. Obsh. Khim. 35 (1965) 443–447.
15] M. Nishio, Cryst. Eng. Commun. 6 (2004) 130–158.
[
2
1
[16] M. Nishio, M. Hirota, Tetrahedron 45 (1989) 7201–7245.
[17] M. Nishio, Y. Umezawa, K. Honda, S. Tsuboyama, H. Suezama, Cryst. Eng.
Commun. 11 (2009) 1757–1788.
[
18] Y. Umezawa, S.T.K. Honda, J. Uzawa, Nishio, Bull. Chem. Soc. Jpn. 71 (1998)
207–1213.
[19] V.T. Nguyen, I.Y.H. Chan, R. Bishop, D.C. Craig, M.L. Scudder, New. J. Chem. 33
2009) 1736–1741.
1
(
(
±) 6,8-Dimethyl-1,3-diphenyl-2,4,10-trioxa-tricyclo[4.4.0.03,8]-
[
[
20] J. Sandström, Dynamic NMR Spectroscopy, London & New York, 1982.
21] Adept Scientific Inc., g-NMR v. 5.0 (Adept Scientific Inc. Bethesda MD), 2003.
1
decane 5: m.p. 178–182 °C; H NMR: 0.64 (6H, s), 1.76 (2H, s),
.56 (2H, d, J = 8.6), 4.01 (2H, d, J = 8.6), 7.35 (6H, m), 7.62 (4H,
m); C NMR: 18.8, 41.1, 41.3, 70.9, 102.4, 126.1, 127.8, 128.3,
37.3. Chiral HPLC analysis: Chiralcel OD-H, Hexane/2-PrOH
99/1), 1 mL/min, 25 °C, UV detection at 254 nm and polarimeter,
3
[22] C. Wolf, Dynamic Stereochemistry of Chiral Compounds: Principles and
Applications, Cambridge, 2008.
1
3
[
23] K.N. Zelenin, V.V. Alekseyev, K. Pihlaja, V.V. Ovcharenko, Russian Chem. Bull.
Int. Ed. Engl. 51 (2002) 205–221.
1
(
[24] F. Macromodel, Mohamadi, N.G.J. Richards, W.C. Guida, R. Liskamp, M. Lipton,
C. Caufield, G. Chang, T. Hendrickson, W.C. Still, J. Comput. Chem. 11 (1990)
Rt(+) = 5.25, Rt(ꢁ) = 5.86, k(+) = 0.72, k(ꢁ) = 0.92,
a = 1.28 and
4
40–467.
25] U. Burkert, N.L. Allinger, in: American Chemical Society, Washington D.C.,
982, MacMimic was available from Instar Software AB, Ideon Research Park,
Rs = 1.01.
[
1
S-223 270 Lund, Sweden.
[
[
26] ADF Program System 2007.01; Scientific Computing & Modelling NV Vrije
Universiteit.
27] G.M. Sheldrick, Acta Cryst. A64 (2008) 112–122.
Acknowledgements
Financial support from the Romanian Ministry of Education
CEx Contract 30/2006) and from CNCSIS–UEFISCSU, Romania
[28] International Tables for Crystallography, Dordrecht, 1992.
29] K. Brandenburg, Bonn, Germany, 2006.
[
(