Diazadioxadecalin and salen podands and macrocycles within dynamic combinatorial virtual libraries: Structure, prototropy, complexation and enantioselective catalysis

Article abstract of DOI:10.1016/S0022-328X(01)00882-8

The reactions of L-1,4-diaminobutanediol (3) and D-2,3-diaminobutanediol (4) with salicyl aldehyde provide the tautomeric manifolds of L-1,4-bis(salicylideneamino)-2,3-butanediol (5) and D-2,3-bis(salicylideneamino)-1,4-butanediol (6), respectively. O-alkylation of the salicyl moiety stabilizes the closed dioxadiazadecalin (DODAD) and diazadioxadecalin (DADOD) isomers (7″, 8″) and accordingly, the dialdehyde 1,2-bis(o-formylphenoxy)-ethane (9) led to the respective macrocyclic manifolds (10-10″ and 11-11″). These tautomeric manifolds are typical target-driven dynamic combinatorial virtual libraries, which can be biased by complex formation with metal ions of different ionic radius. A rare instance of simultaneous occurrence of keto-enamine and phenol-imine tautomers in the solid state of 6 was unravelled (X-ray at two temperatures) and the strength of the intramolecular hydrogen bonding (and hence, the extent of ring closure) in 6 is temperature dependent. Compounds 6, 11 and 12-14 constitute a new class of salens, which form heavy and transition metal complexes. Some such Mn(III) complexes are good chirality inducing catalysts, as found in asymmetric indene epoxidation reactions.

Full text of DOI:10.1016/S0022-328X(01)00882-8

Journal of Organometallic Chemistry 630 (2001) 67–77
www.elsevier.com/locate/jorganchem
Diazadioxadecalin and salen podands and macrocycles within
dynamic combinatorial virtual libraries: structure, prototropy,
complexation and enantioselective catalysis
Alexander Star, Israel Goldberg, Benzion Fuchs *
School of Chemistry, Tel-A6i6 Uni6ersity, Ramat-A6i6, 69978 Tel A6i6, Israel
Received 8 January 2001; received in revised form 20 March 2001; accepted 22 March 2001
Dedicated to Professor Myron Rosenblum on the occasion of his 75th birthday
Abstract
The reactions of
manifolds of -1,4-bis(salicylideneamino)-2,3-butanediol (5) and
O-alkylation of the salicyl moiety stabilizes the closed dioxadiazadecalin (DODAD) and diazadioxadecalin (DADOD) isomers (7¦,
L
-1,4-diaminobutanediol (3) and
D
-2,3-diaminobutanediol (4) with salicyl aldehyde provide the tautomeric
L
D
-2,3-bis(salicylideneamino)-1,4-butanediol (6), respectively.
8
1
¦) and accordingly, the dialdehyde 1,2-bis(o-formylphenoxy)-ethane (9) led to the respective macrocyclic manifolds (10–10¦ and
1–11¦). These tautomeric manifolds are typical target-driven dynamic combinatorial 6irtual libraries, which can be biased by
complex formation with metal ions of different ionic radius. A rare instance of simultaneous occurrence of keto–enamine and
phenol–imine tautomers in the solid state of 6 was unravelled (X-ray at two temperatures) and the strength of the intramolecular
hydrogen bonding (and hence, the extent of ring closure) in 6 is temperature dependent. Compounds 6, 11 and 12–14 constitute
a new class of salens, which form heavy and transition metal complexes. Some such Mn(III) complexes are good chirality inducing
catalysts, as found in asymmetric indene epoxidation reactions. © 2001 Elsevier Science B.V. All rights reserved.
Keywords: Salens; Dioxadiazadecalins; Tautomerism; Combinatorial virtual libraries; Complexes; Enantioselective catalysis
1
. Introduction
|+ is the Hammett–Brown constant of the 2(6)-aryl
substituents) and the slope z was 0.69, close to that
found earlier in ring–chain tautomeric studies [2] of
1,3-oxazanes [2a].
We have recently put forward the novel cis- and
trans-1,5-dioxa-3,7-diazadecalin (DODAD) and -1,5-di-
aza-3,7-dioxadecalin (DADOD) systems (Scheme 1) [1],
with various para-aryl substituents and their interesting
tautomeric equilibrium with the respective Schiff bases
(
1 X 1¦ and 2 X 2¦) [1c].
cis-DODAD (cis-1¦) systems prevailed only with
electron-withdrawing substituents, but cis-DADOD
(
cis-2¦) systems were more stable than their parent
Schiff bases in all cases. The ring–chain tautomeric
process has been analyzed [1c] using the Hammett
+
equation (log K=z| +log K , K=[ring]/[chain] and
0
*
Corresponding author. Tel.: +972-3-640-8623; fax: +972-3-640-
293.
E-mail address: bfuchs@post.tau.ac.il (B. Fuchs).
Scheme 1. The 1,4-bis(arylideneamino)-2,3-butanediol (1)/dioxadi-
azadecalin (DODAD, 1¦) and 2,3-bis(arylideneamino)-1,4-butanediol
(2)/diazadioxadecalin (DADOD, 2¦) equilibria.
9
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(01)00938-X

68
A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
Double Schiff bases in open or cyclic array and, in
particular the salen systems are, as long known, good
ligands [10,11]. Chiral salen manganese complexes
based on optically active 1,2-diamines (mainly 1,2-di-
aminocyclohexane) have been used as synthetically use-
ful catalysts for the asymmetric epoxidation of a variety
of olefins [11]. Complexes with other metals have been
also prepared and their structures studied in the solid
state [10,11]. It is on this background, that we report
now our findings in this area.
Scheme 2. Phenol–imine/keto–enamine tautomerism.
The hydroxy and alkoxy groups have similar induc-
tive effect (| =0.12), but contribute by resonance
m
+
rather differently (| = −0.92 vs. −0.78) [3]. The OH
group is more electron-donating than OCH , due to the
3
contribution of the keto-resonance form. All linear free
energy relationship studies have dealt only with meta
and para substituents, because ortho substituted 2-aryl
systems are outstanding [2c], often exhibiting hydrogen
bonding and steric hindrance in addition to electronic
effects. We were, however, very much interested in
salicyl derivatives, as precursors for macrocylic systems
and complexation agents [4].
N-Salicylideneaniline derivatives have been shown to
exist as tautomeric forms due to intramolecular proton
shifts between the (phenol) oxygen and the (imine)
nitrogen, via intramolecular hydrogen bonding
2
. Results and discussion
The reaction of salicyl aldehyde with
L-1,4-di-
-1,4-bis(salicylide-
aminobutanediol (3) provided
neamino)-2,3-butanediol (5), while
butanediol (4) gave -2,3-bis(salicylideneamino)-1,4-bu-
L
D
-2,3-diamino-
D
tanediol (6) (Scheme 3). These Schiff bases had been
prepared earlier [12] and we report now their full
physical characterization. Moreover, no attention had
been paid earlier to any possible tautomeric forms and,
in fact, we observed (by NMR monitoring) formation
of the isomeric dioxadiazadecalin systems, albeit with
prevalence and final crystallization of the open forms.
The stability of the latter is undoubtedly due to strong
intramolecular OꢀH···O and OꢀH···N hydrogen bond-
ing. 5 is all but insoluble in chloroform and its NMR
spectrum in DMSO-d6 shows no closed tautomeric
forms. Similarly, in a variable solvent and temperature
(NMR) probe of 6, no common solvent was able
to overcome the intramolecular hydrogen bonding
in the open tautomer (6) and to stabilize the
closed forms (6% and 6¦) (Table 1). However, on lower-
(
OꢀH···N X O···HꢀN) (Scheme 2, R=aryl) [5–9].
These prototropic shifts, which occur also in the well-
known salen systems, are inevitably accompanied by
p-electron configurational changes, either in the ground
state or in the excited state, in solution and in the
crystal. Some crystals of such internally hydrogen
bonded Schiff bases exhibit thermochromism by a ther-
mally induced intramolecular proton shift to the
quinoid tautomer, or photochromism, by an irradiation
induced proton migration [5a,5d].
Scheme 3. The formation of the 5–5¦ (7–7¦) and 6–6¦ (8–8¦) tautomeric manifolds in the reactions of salicylaldehyde (o-anisylaldehyde) with
-1,4-diamino-2,3-butanediol (3) or -2,3-diamino-1,4-butanediol (4), respectively.
L
D

A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
69
Table 1
Some of the stability of the bicyclic forms may be
attributed to NꢀH···OR hydrogen bonding, but this is
somewhat weakened by the steric interference of two
OMe groups. This interference should abate by connect-
ing the two phenolic termini, leading to the much desired
macrocyclic DODAD and DADOD systems. The length
of the best fitting spacer was probed by MM calculations
(CFF91) of different glycol spacers in a DODAD model,
the criterion being the phenolic O···O distance, which is
Selected NMR data and composition vs. temperature and solvent
(
from NMR) for 6–6¦, 7–7¦, 10 and 12
X
T (K)
X (%)
X% (%)
X¦ (%)
l
l
NꢁCH
PhꢀOH
a
6
6
7
223
70
100
100
100
16
25
5
13.52
13.06
13.54
13.11
8.46
8.40
8.54
8.52
2
98
298
23
298
23
298
23
298
b
3
a
30
27
54
42
100
100
95
85.5
40
8
2.70–3.07 A
,
in an unsubstituted strainless molecule (e.g.
3
39
5
ꢀ
¦). The explored bridges were: ꢀOCH Oꢀ (2.34 A),
,
2
a
1
1
0
OCH CH Oꢀ (2.88 A
,
), ꢀOCH CH CH Oꢀ (3.52 A
,
)
2
2
2
2
2
3
b
and
ꢀOCH
ꢀOCH CH OCH CH Oꢀ
(5.00
A
,
).
For
0
3.5
9.5
28
45
100
1.5
5
32
47
2
2
2
2
3
3
4
23
68
10
CH
2
2
Oꢀ this distance roughly coincided with that
of the unsubstituted molecule.
Our first attempts of O-alkylation on the Schiff bases
was largely ineffectual, but using the known dialdehyde
1,2-bis(o-formylphenoxy)-ethane (9) [10a] in the reaction
with each of the diaminobutanediols (Scheme 4) at high
dilution, readily provided the desired macrocycles,
(2R,6R;9S;9,10-P)-2,6-(1%,2%;7%,8%-dibenzo-3%,6%-dioxa-
a
12
298
13.30
8.39
a
Solvent: CDCl3.
Solvent: DMSO-d6 (also MeOH-d4 and Py-d5).
b
ing the temperature (to −50 °C in CDCl ), both closed
oxazane forms (prevalently 6%) were observed.
3
1%,8%-octanylidene)-cis-dioxadiazadecalin
(10¦)
and
As in simpler 2-hydroxy systems [3], these internal
hydrogen bonds could be averted by oxygen alkylation.
Indeed, reaction of 3 or 4 with o-anisaldehyde (Scheme
(2R,6R;9S;9,10-M)-2,6-(1%,2%;7%,8%-dibenzo-3%,6%-dioxa-
1%,8%-octanylidene)-cis-diazadioxadecalin (11¦) in all but
quantitative yields and, notably, with no help of any
metal template. The structure of 10¦ has been unequiv-
ocally proven by X-ray diffraction analysis [4].
Strong intramolecular hydrogen bonding of NH···OR
type in 10¦ and 11¦ is manifest in each by a clear double
doublet for NH with large anti vicinal coupling con-
stants, and by strong deshielding (2 ppm) of these
protons, compared to the earlier p-nitrophenyl substi-
tuted molecules [1c,1d]. The X-ray structure of 10¦
3) produced the 7 and 8 O-methylated molecular mani-
folds with higher dioxadiazadecalin stability. In case of
DODAD (7), the ratio is: 7/7%/7¦=16/30/54% (Table 1)
and the contribution of ring closed species (7%,7¦) dimin-
ished at higher temperature. However, in the case of 8,
there is complete ring closure to the DADOD form (8¦).
It should be reiterated that the formation and all shown
reactions of 5 (7) and 6 (8) are completely stereospecific.
Scheme 4. The formation of the 10–10¦ and 11–11¦ tautomeric macrocyclic manifolds.

70
A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
mono- and di-imine forms (10 and 10%) becomes signifi-
cant in DMSO-d , in particular at higher temperatures
6
(
Table 1).
We regard these manifolds (Schemes 3 and 4) as
albeit small) Dynamic Combinatorial Virtual Library
(
[4]. These are sets of isomeric systems, any member of
which may but need not exist in significant amounts, if
not in the presence of a recognizable target or de-
manded by extraneous conditions (medium, tempera-
ture, etc.) [13]. Our above ring–chain tautomeric
molecular sets, as well as analogous ones previously
described [1c,1d,4], qualify particularly well for this
definition.
Fig. 1. ORTEP drawing of 2,3-bis(salicyldeneamino)-1,4-butanediol
6).
(
We turn to the interesting behavior of the salen type
compound (6) (Scheme 3) and its corresponding
DADOD (6¦) derivative and their complexes.
confirms these strong intramolecular hydrogen bonds
4], which are bound to contribute considerably to the
[
The crystal structure of
D-2,3-bis(salicylideneamino)-
smooth macrocyclization reaction.
1
,4-butanediol (6) was determined by X-ray diffraction
Each of these macrocyclic dioxadiazadecalins (10¦,
1¦) is actually a member of a manifold of three
at both 295 and 116 K (Fig. 1 and Table 2(a)). There is
evidently strong intramolecular ArꢀOH···N hydrogen
bonding in one wing, but in the second one the hydro-
gen is located mainly at the nitrogen with strong hydro-
gen bonding to the oxygen site. Cooling enhanced this
effect, which is a clear case of phenol–imine/keto–
1
possible tautomeric structures (Scheme 4) (without
mentioning the respective five-membered higher energy
isomeric forms) [4]. Indeed, while 10¦ is exclusive in
CDCl3 even when heated, the contribution of the
Table 2
Relevant structural parameters from X-ray analysis (a) of 6 at two temperatures and (b) of selected literature examples of salens and salophenes
Fig. 2), retrieved from the Cambridge Structural Database (CSD) [14]
(
Interatomic distances (A
,
)
Dihedral angles (°)
Compound d1
d2
d3
d4
d5
d6
ꢀ a
„b
(a)
6295 K
1.348
.320
1.344
.293
1.396
1.407
1.416
1.440
1.453
1.431
1.454
1.429
1.269
1.295
1.281
1.297
2.624
2.535
2.666
2.556
2.927
2.866
53.9
52.6
15.7
16.1
1
6116 K
b)
1
(
I [7a]
II [7b]
1.345
1.346
1.412
1.425
1.415
1.397
1.401
1.406
1.416
1.407
1.412
1.414
1.403
1.429
1.457
1.442
1.456
1.452
1.446
1.452
1.447
1.471
1.467
1.462
1.451
1.406
1.270
1.292
1.284
1.283
1.285
1.270
1.276
1.284
1.288
1.271
1.272
1.303
2.597
2.627
2.570
2.620
2.612
2.604
2.610
2.521
2.508
2.604
2.602
2.578
3.523
3.517
6.082
5.544
5.868
8.2
5.3
60.0
45.0
56.5
83.1
76.3
1
.345
1.358
.358
1.348
.347
1.346
.340
1.365
.361
1.277
III [8a]
IV [8b]
V [8c]
1
58.6
63.1
64.9
1
1
VI [8d]
1
VII [8e]
a
NꢀCꢀCꢀN torsion angle.
Dihedral angle between the two aromatic rings
b

A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
71
enamine (N···HꢀOꢀCDNꢀH···OꢁC) tautomerism. From
comparing the structural parameters of the hydrogen
bonded chelate ring at both temperatures, the stronger
OH···N hydrogen bond was observed at room tempera-
ture, as indicated by the shorter N···O distance.
The Cambridge Structural Database (CSD) [14] con-
tains numerous salen type complexes but only 12 struc-
tures of free ligands. The relevant geometrical
parameters of the hydrogen-bonded chelate rings of
some of these salen and salophen type compounds
While some of the existing structures show slight lack
of symmetry between the two aromatic wings (Table
2(b)), it can be seen from the geometrical parameters of
the hydrogen-bonded chelate rings that only in our case
(6) there is complete asymmetry, due to the two differ-
ent, phenol–imine and keto–enamine, tautomeric
flanks (cf. VII for typical data of keto–enamine spe-
cies). To the best of our knowledge, 6 is the first
example of coexistence of two tautomers in the same
salicylaldimine molecule. The very short O···O distance
and near coplanarity (16°) of the aromatic rings (Table
2(a)) (and implicitly of the two pseudoaromatic chelate
rings) are outstanding features in this context.
(
I–VI) (Fig. 2) are assembled in Table 2(b), for com-
parison with our own results.
The pseudoaromatic six-membered ring formed by
intramolecular hydrogen bonding in general N-salicyli-
deneanilines and naphtylamines is planar and ranges
between non-chromatic to photochromic. The hydro-
gen bond type appears to depend neither on the stereo-
chemistry of the molecule nor on the RꢀNꢁC
substituent, but on the kind of aldehyde used [9] and, in
solid state, the keto–enamine tautomer (cf. Scheme 2)
appears to dominate in naphthaldimine (e.g. VII), while
the phenol–imine prevails in salicylaldimine Schiff
bases. Thus, the existence of the enol tautomer has been
established in all crystal structures of N-alkyl substi-
tuted salicylaldimines listed so far in the Cambridge
Structural Database and an intramolecular NꢀH···O
hydrogen bond is present in naphthaldimines in the
solid state regardless of the kind of N-substituent, aryl
or alkyl. Indeed, the salen (I, with is NꢀCꢀCꢀN anti
conformation) [7a] and the salophen (II, III) [7b]
derivatives (Table 2(b)) exhibit bond lengths and angles
consistent with double NꢁC and single OꢀC bonds, in
full conformity with the enoliminic structure.
In solution, UV–vis spectroscopic studies showed
that the keto–enamine form exhibits an absorption
band at \400 nm, always observed in 2-hydroxynaph-
thaldehyde/aniline Schiff bases, but not in salicylalde-
hyde/aniline ones (except in acidic media). The
extinction coefficients of the keto–enamine bands at
\400 nm depend on solvent polarity, decreasing from
EtOH, via DMSO and CHCl , to cyclohexane [9].
3
NMR spectra also reflect the intramolecular NꢀH···O
hydrogen bonds in Schiff bases of the N-salicylide-
neaniline and N-naphthylideneaniline type in solution
1
[9]. The H-NMR signals for the phenol–imine form in
CDCl are singlets at l 13.75 ppm for ArOH and at l
3
8.73 ppm for CHꢁN, whereas the keto–enamine form
in the same solvent peaks at l 15.57 ppm and at l 9.20
3
ppm ( J=7.6 Hz).
The corresponding data of 6 (Table 1) indicate that
cooling in CDCl increase the keto–enamine contribu-
3
tion, whereas heating in DMSO enhance the phenol–
imine form and the variable solvent NMR data show
Fig. 2. Literature examples of salens and salophenes, retrieved from the Cambridge Structural Database (CSD) [14].

72
A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
that the keto–enamine form is enhanced as function of
solvent polarity. This was, in fact, corroborated by
UV–vis spectral data of both 6 and 12 (see Section 4),
which showed an enhanced contribution of the keto–
amine forms (at \400 nm) in MeOH, as compared to
MeCN.
A compelling sequel was to probe the complexation
ability of these systems. The described Dynamic Com-
binatorial Virtual Libraries (DCVL) (Schemes 3 and 4)
are far from being ideally unbiased and, in principle,
one could scan them for their constituents, by varying
temperature and/or solvent (Table 1). However, any
biased or unbiased DCVL, when target-driven, the
preferred interaction of a virtual or insignificant con-
stituent with the target may overturn the equilibrium in
the set. An obvious choice of target is metal ion
complexation.
chloride in methanol, the end product is exclusively
2
+
(NMR) the dioxadiazadecalin complex (10¦·Ni
)
(Scheme 5). In the FABMS experiment, however, the
+
occurrence of peaks at 447 (30%) of [NiLCl] and at
+
411 (65%) of [NiL−H] , indicate also the presence of
2
+
the tautomeric diimine complex (10·Ni ). This is a
general and diagnostic feature in all FABMS analyses
of complexes in the macrocyclic series 10 and 11: the
complexes of both the bicyclic DODAD (10¦) and
+
DADOD (11¦) exhibit [MLX] ions, but only in the
ring–chain tautomeric 10 series can one find an [ML−
+
H] ion characteristic for the dimine (10).
When either cadmium perchlorate or lead perchlorate
were added to 10¦ in methanol, the respective com-
plexes precipitated, but could, however, be dissolved in
DMSO-d for NMR purposes. It turned out that during
6
2
+
complexation with Cd , 10¦ underwent tautomeric
ring opening and all forms (10, 10%, 10¦) became present
in the complexes. This was corroborated by FABMS
In the reaction of 11¦ with nickel chloride in
methanol at room temperature, we observed (NMR)
the slow formation of a transient imine form (at 8.5
ppm), but the final product was clearly a DADOD
complex and (Scheme 5), if a diimine macrocyclic com-
+
(peaks at 567 (25%) [CdL(ClO4)] and 467 (7%)
+
+
[CdL−H] and 465 (15%) [CdL−3H] ). In the case
of lead, however, the NMR spectrum in DMSO-d
shows only the diimine complex (10·Pb ) (Scheme 5),
6
2
+
2+
plex 11·Ni
had initially formed [10a], it underwent
efficient tautomeric ring closure to the more stable
with FABMS peaks characteristic to the open chain
2
+
+
complex 11¦·Ni . The latter’s FABMS analysis reveals
form (561 (70%) [PbL−H] and 559 (140%) [PbL−
+
+
2+
2+
a peak at 447 (20%) that fits exactly the [NiLCl]
3H] ). The larger Cd
(1.18 A) ions appear to be readily complexed by 10¦ (in
(0.97 A) and the largest Pb
,
species. Similarly, complexation of 11¦ with Ni(ClO )
,
4
2
2
+
in acetonitrile afforded 11¦·Ni , as observed by UV–
vis spectrometry, by which a 1:1 complexation stoi-
chiometry could also be deduced [4].
contrast to 11¦) due to its ability (Scheme 2) to modify
its cavity size by a successive ring-opening process,
ending up with a ring increase by two atoms to
2
+
In contrast, attempts to isolate or observe a lead ion
10·Pb . Eventually, the latter may undergo gradual
reverse macrocycle folding, providing a mechanism of
fine-tuning until an optimum cavity size for ion bond-
ing is reached. That is how Ni(II), Cd(II) and Pb(II)
were complexed by 10–10¦ in different, cation size
depending tautomeric modes.
2
+
complex 11¦·Pb
in methanol failed, apparently be-
ion of small ionic radius (0.69 A) can
readily inhabit its DADOD cavity, which the larger
2
+
cause the Ni
,
2
+
metal ion Pb
(1.18 A) cannot (Scheme 5). The tau-
,
tomeric ring opening within the macrocylic ligand to
the diimine (11) doesn’t really help, since its cavity is
similarly sized.
In the dioxadiazadecalin (DODAD) macrocyclic se-
ries (Scheme 2, top), after treatment of 10¦ with nickel
Having shown that cyclic diazadioxadecalin/salen
systems complex well with selectively heavy and transi-
tion metal ions, it was compelling then to examine the
complexation ability of 6 (vide supra) and of judi-
ciously chosen related salen analogues and the catalytic
activity [11,15] of the complexes. We picked the
Mn(III) complexes, in particular since we could use
Jacobsen’s well-forged procedures [15,16] for their use
as catalysts of epoxidation of olefins and indene as a
prototypical substrate [16]. Thus, after preparing 6
(
Schemes 3 and 5) and its dibenzyl ether (12) [1d] and
their Mn(III) complexes (15, 16) (Scheme 6), and realiz-
ing, as actually anticipated [11], that the latter catalyse
the epoxidation of indene but without any asymmetric
induction, we prepared the tetra-tert-butyl derivatives
1
3 and 14 and their Mn(III) complexes (17 and 18). The
use of the latter in the epoxidation of indene (Scheme 7)
gave 2,3-epoxyindane in yields (ee) of 50% (48%) and
5
1% (30%), respectively. This is less then the high yield
Scheme 5. Host–guest size interdependent complex formation in the
10–10¦ and 11–11¦ tautomeric macrocyclic series.
achieved by Jacobsen [11,15,16b] with his rigid salen

A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
73
Scheme 6. The synthesis of
D
-2,3-bis(arylideneamino)-1,4-butanediol derivatives (12–14) and of their Mn(III) complexes (15–18).
lysts, as established in moderately enantioselective in-
dene epoxidation reactions.
4
. Experimental
4.1. General
Melting points were recorded on a capillary melting
point apparatus and are uncorrected. The H- and
C-NMR spectra were obtained on 200 or 500 MHz
Scheme 7. The epoxidation of indene using the Mn(III) complexes of
-2,3-bis(di-t-butyl-salicylideneamino)-1,4-butanediol and -dibenzyl
ether (17, 18).
1
D
13
spectrometers and referenced to Me Si as internal stan-
4
dard. Mass spectra (DEI-MS, DCI-MS and FAB) were
recorded on an Autospec 250 mass spectrometer. UV
spectra were taken on a Unikon 931 spectrophotome-
ter. Since we dealt with tautomeric mixtures, optical
purity was measured using chiral NMR shift reagents,
built on the trans-cyclohexano core, but sufficient to
maintain our incentive for further probing and optimiz-
ing these readily available and variegated catalytic
systems.
usually Eu(hfc) , unless otherwise mentioned.
3
3. Conclusions
4.2. General procedures
We have shown that
L
-1,4-bis(salicylideneamino)-2,3-
L
-1,4-Diaminobutanediol (3) and
D-2,3-diamino-1,4-
butanediol (5) and
D
-2,3-bis(salicylideneamino)-1,4-bu-
butanediol (4) were prepared as described [1d]. 3,5-Di-
t-butylsalicylaldehyde was prepared by a published
procedure [16a]. Each diaminobutanediols or derivative
was stirred at room temperature with two mole equiva-
lents of the chosen salicylaldehyde in EtOH, until full
conversion.
tanediol (6) exist in solvent and temperature dependent
tautomeric equilibria, with the ring-closed (DODAD
and DADOD) systems. The equilibrium is influenced
by phenol O-alkylation, stabilizing the closed dioxadi-
azadecalin
(DODAD)
and
diazadioxadecalin
(
DADOD) isomers (7¦, 8¦). This led the way to using
L
-1,4-Bis(salicylideneamino)-2,3-butanediol (5) was
the dialdehyde 1,2-bis(o-formylphenoxy)-ethane (9) for
securing the respective macrocyclic manifolds (10–10¦
and 11–11¦). These tautomeric manifolds are typical
target-driven dynamic combinatorial 6irtual libraries,
which can be biased by complex formation with metal
ions of different ionic radius. We unravelled a unique
instance of simultaneous occurrence of keto–enamine
and phenol–imine tautomers in solution and in the
solid state of 6. The latter and a number of analogues
constitute a new class of salens, which form heavy and
transition metal complexes. Such Mn(III) complexes
prepared from -1,4-diamino-2,3-butanediol [12c] (3)
and salicylaldehyde. 5 precipitated from EtOH solution
as yellow crystals; m.p. 224 °C (MeOH). H-NMR
(DMSO-d ): l=13.67 (brs, ArOH), 8.55 (s, NꢁCH),
7.45 (dd, J=7.6, J=1.0 Hz, H ), 7.35 (dt, J=7.8,
J=1.3 Hz, H ), 6.90 (t, J=7.5 Hz, H ), 6.89 (d,
J=8.0 Hz, H ), 5.02 (brd, J=4.6 Hz, OH), 3.89 (dd,
J=12, J=3.3 Hz, H ), 3.77 (m, H ), 3.58 (dd, J=
12, J=7.6 Hz, H ). C-NMR (DMSO-d ): l=166.9
(d, NꢁCH), 161.1, 118.7 (s, Ar), 132.3, 131.7, 118.4,
116.6 (d, Ar), 71.5 (d, C ), 61.2 (t, C ). DCIMS; m/z
(relative intensity %): 329.1 ([MH] , 100). UV: u
L
1
6
3
4
3
a
4
3
2
3
c
b
3
d
3
2
1
2
3
13
1
6
2
1
+
(
17, 18) are potentially good chirality inducing cata-
max

74
A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
−
1
−1
(
2
nm) (m M cm ) (MeOH): 402 (3000), 315 (6800),
55 (22 200), 216 (43 200).
-2,3-Bis(salicylideneamino)-1,4-butanediol (6) was
prepared from -2,3-diamino-1,4-butanediol [12c] (4)
EtOH. The solution was stirred at room temperature
(r.t.) until complete conversion (TLC, SiO , CHCl ).
Subsequent evaporation provided white solid products.
2
3
D
1
D
10¦: m.p. 172 °C (MeCN). H-NMR (CDCl ): l=
3
3
4
3
and salicylaldehyde. 6 precipitated from EtOH solution
as yellow crystals, m.p. 163 °C (MeOH). H-NMR
CDCl ): l=13.06 (brs, OH), 8.40 (s, NꢁCH), 7.31 (dt,
J=7.8, J=1.6 Hz, H ), 7.24 (dd, J=7.7, J=1.5
Hz, H ), 6.98 (d, J=8.3 Hz, H ), 6.86 (dt, J=7.5,
J=0.8 Hz, H ), 3.90 (dd, J=11.2, J=4.1 Hz, H ),
7.31 (dt, J=7.9, J=1.7 Hz, H ), 7.26 (dd, J=7.8,
c
1
4
3
3
a d
J=1.7 Hz, H ), 7.07 (d, J=7.8 Hz, H ), 6.98 (dt,
4
3
(
J=7.4, J=0.8 Hz, H ), 5.19 (d, J=12.3 Hz, H ),
b 2
4.54, 4.31 (ddd, ꢀCH CH ꢀ), 3.94 (t, J=12.2 Hz,
NH), 3.61 (brs, H ), 3.28 (dd, J=14.5, J=1.5 Hz,
3
3
4
3
4
3
c
2
2
3
3
2
3
a
d
9
4
2
3
13
H4eq), 3.10 (t, J =13 Hz, H4ax). C-NMR (CDCl3):
b
1
av
2
3
3
(
1
.83 (dd, J=11.0, J=6.7 Hz, H ), 6.67 (m, H ), 1.88
l=130.3, 129.8, 121.8, 114.1 (d, Ar), 92.1 (d, C ), 69.6
1
2
2
1
3
brs, OH). C-NMR (CDCl ): l=168.2 (d, NꢁCH),
(d, C ), 67.2 (t, C
), 50.3 (t, C ). CIMS; m/z (relative
3
9
bridge 4
+
33.6, 132.5, 119.6, 117.9 (d, Ar), 72.6 (t, C ), 64.4 (d,
intensity %): 355.2 ([MH] , 100), 311.1 (14), 253.1 (9).
1
+
+
’
HR DEIMS: Calc. for C H N O [M]
C ). DCIMS; m/z (relative intensity %): 329.1 ([MH] ,
00). UV–vis: umax (nm) (m, M cm ) (MeOH): 403
354.1580,
observed m/z 354.1574. [h] = +32° (c=5.2, CHCl ).
2
20 22
2
4
−
1
−1
1
D
3
(
1000), 315 (7500), 256 (23 000), 214 (40 500); (MeCN):
The other two tautomers could be seen in the above
4
02 (95), 315 (7844), 256 (22 434), 215 (41 935).
The reaction of 3 and o-anisaldehyde in ethanol gave
crude product, in DMSO at higher temperature: 10:
1
H-NMR (DMSO-d , 410 K): l=8.76 (s, NꢁCH), 7.81
6
3
4
3
4
a tautomeric mixture of 2,6-bis(o-anisyl)-cis-DODAD
7¦) together with small amounts of 7, 7%. On evapora-
tion a colorless solid was obtained, m.p. 120–129 °C
(dd, J=7.6, J=1.5 Hz, H), 7.41 (dt, J=7.2, J=
3
3
(
1.7 Hz, H), 7.15 (d, J=8.0 Hz, H), 7.02 (dt, J=7.5,
4
J=2.0 Hz, H), 4.48 (m, ꢀCH CH ꢀ), 3.85 (m, H ),
2
2
2
1
1
(
t-BuOMe). H-NMR (CDCl ): (7) l=8.76 (s, NꢁCH);
3.79 (m, 2H ). 10%: H-NMR (DMSO-d , 410 K): l=
3
1
6
3
3
4
(
7%) 8.68 (s, NꢁCH), 5.60 (s, H ); (7¦) 7.64 (dd, J=7.5,
8.70 (s, NꢁCH), 7.89 (dd, J=7.5, J=1.6 Hz, H), 7.42
3 3
2
4
3
4
J=1.4 Hz, H ), 7.34 (dt, J=7.8, J=1.6 Hz, H ),
(overlap, H), 7.34 (t, J=8.0 Hz, H), 7.25 (dd, J=7.4,
a
c
3
3
4
3
7
.05 (t, J=7.6 Hz, H ), 6.95 (d, J=8.6 Hz, H ), 5.65
J=1.5 Hz, H), 7.20 (d, J=8.0 Hz, H), 7.15 (overlap,
b
d
2
3
(
s, H ), 3.89 (s, CH ), 3.68 (s, H ), 3.35 (dd, J=14.5
H), 7.02 (overlap, H), 6.98 (t, J=7.4 Hz, H) 5.15 (s,
H ), 4.60, 4.48, 4.24 (m, ꢀCH CH ꢀ), 3.63 (m, 2H ),
2
3
9
2
13
Hz, H ), 3.27 (d, J=14.5 Hz, H ). C-NMR (CDCl3):
4
4
2
2
2
7
(
1
7) l=159.6 (d, NꢁCH), 72.4 (d, C ), 65.3 (t, C ); (7%)
3.58 (m, H +H ), 3.04 (m, H ).
2
1
5 6 4
1
56.4 (d, NꢁCH), 84.9 (d, C ), 66.1 (d, C ), 64.6 (t,
Similarly, 11¦ m.p. 127 °C (MeCN). H-NMR
3 4
2
5,6
C ), 51.0 (t, C ); (7¦) 129.6, 126.9, 120.8, 111.0 (d, Ar),
8
(CDCl ): l=7.32 (dt, J=7.9, J=1.7 Hz, H ), 7.28
3 c
7
4
3
4
3
4.6 (d, C ), 69.9 (d, C ), 55.6 (OCH ), 50.1 (t, C ).
(dd, J=7.4, J=1.7 Hz, H ), 7.08 (d, J=7.8 Hz,
H ), 7.00 (dt, J=7.4, J=0.8 Hz, H ), 5.19 (d, J=
13.2 Hz, H2ax), 4.59, 4.38 (ddd, H
2
9
3
4
a
−
1
−1
3
4
3
UV–vis: umax (nm) (m, M cm ) (MeOH): 310
6000), 252 (17 600), 199 (91 600); (MeCN): 305 (8409),
51 (20 272), 212 (33 182). DCIMS; m/z (relative inten-
d
b
3
bridge av
(
2
), 4.49 (t, J =
2
3
12.5 Hz, NH), 4.14 (dd, J=11.6, J=1.3 Hz, H4eq),
+
2
3
3
sity %): 357.3 ([MH] , 100), 249.2 (7), 239.2 (9), 221.2
4.05 (dd, J=11.7, J=1.8 Hz, H ), 2.95 (d, J=11.2
4ax
1
3
(
(
55), 208.2 (55), 178.1 (79), 149.1 (28), 136.1 (21), 121.1
35), 104.1 (9), 91.1 (13). HR DCIMS; m/z: 357.1816.
Hz, H9). C-NMR (CDCl ): l=156.6, 128.2 (s, Ar),
3
130.6, 130.0, 122.1, 114.5 (d, Ar), 91.9 (d, C ), 71.7 (t,
2
+
Calc. for C H N O [MH] 357.1814.
C ), 50.1 (d, C ), 66.9 (t, Cbridge). DCIMS; m/z (relative
2
0
25
2
4
4
9
+
The reaction of 4 with o-anisaldehyde in EtOH gave
on evaporation 2,6-bis(o-anisyl)-cis-DADOD (8¦), as a
intensity %): 355.1 ([MH] , 100), 324.1 (77), 253.1 (60),
121.0 (30). HR MS: Calc. for C H N O [M]
+
’
20
22
2
4
1
3
yellow oil. H-NMR (CDCl ): l=7.50 (dd, J=7.6,
354.1580; observed m/z 354.1576. [h] = −85° (c=2,
3
D
4
3
4
J=1.6 Hz, H ), 7.32 (dt, J=7.8, J=1.7 Hz, H ), 7.0
CHCl3).
a
c
3
(
m, H ), 6.94 (d, J=8.3 Hz, H ), 5.49 (s, H ), 4.17 (d,
J=11.6 Hz, H ), 4.07 (d, J=11.7 Hz, H ), 3.89 (s,
D
- 2,3 - Bis(salicylideneamino) - 1,4 - benzyloxybutane
b
d
2
2
2
(12) was prepared from -2,3-diamino-1,4-benzy-
D
4
4
1
3
OCH ), 3.02 (s, H ). C-NMR (CDCl ): l=129.8,
1
loxybutane [12c] (4a) and salicyladehyde. Evaporation
3
9
3
1
27.4, 120.8, 111.4 (d, Ar), 86.1 (d, C ), 71.5 (t, C ),
gave a yellow oil product. H-NMR (CDCl ): l=13.30
2
4
3
5
5.8 (q, OCH ), 49.9 (d, C ). DCIMS; m/z (relative
(brs, OH), 8.39 (s, NꢁCH), 7.26 (m, Ha,c+Ph), 7.02 (d,
3
9
+
3
2
3
d b
intensity %): 357.3 ([MH] , 100), 326.3 (8), 249.2 (18),
J=8 Hz, H ), 6.88 (t, J=7.3 Hz, H ), 4.50 (d,
2
2
1
39.2 (10), 223.2 (26), 206.2 (17), 192.2 (37), 179.2 (71),
62.2 (29), 136.1 (91), 121.1 (56), 91.1 (29). HR
J=11.9 Hz, CHHPh), 4.44 (d, J=11.9 Hz,
2
3
CHHPh), 3.85 (m, H ), 3.70 (dd, J=9.4, J=5.3 Hz,
2
+
2
3
13
DCIMS; m/z: 357.1816. Calc. for C H N O [MH]
H ), 3.52 (dd, J=9.4, J=6.8 Hz, H1). C-NMR
2
0
25
2
4
1
3
57.1814.
(CDCl ): l=167.2 (d, NꢁCH), 161.4 (s, Ar), 132.6,
3
The macrocyclic compounds 10¦ and 11¦ were pre-
131.6, 128.4, 127.7, 118.6, 117.3 (d, Ar), 73.4 (t, C1),
pared at high dilution: a 0.1 M ethanolic solutions of
diaminobutanediol 3 or 4 (1 mmol) was added to a
solution of the dialdehyde 9 (1 mmol) in 400 ml of
70.8 (t, CH Ph), 69.1 (s, C ). UV–vis: u
(nm) (m,
2
2
max
−
1
−1
M
cm ) (MeOH): 406 (1060), 316 (9200), 257
(30 400); (MeCN): 403 (90), 317 (9000), 257 (26 240),

A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
75
3
3
2
13 (56 164). DCIMS; m/z (relative intensity %): 509.3
7.2 Hz, H ), 7.10 (d, J=6.7 Hz, H ), 7.01 (d, J=7.8
c
a
+
3
d b 2
([MH] , 100), 405.3 (18), 255.2 (8), 122.1 (21), 106.9
Hz, H ), 6.80 (t, J=6.8 Hz, H ), 5.09 (s, H ), 4.38,
(
42), 91.0 (70). HR DCIMS; m/z: 509.2439. Calc. for
4.07 (brd, J=6.5, ꢀCH CH ꢀ), 3.65 (brs, H ), 3.06 (m,
2
2
9
+
C H N O [MH] 509.2440.
H
4
). FABMS; m/z (relative +intensity %): 447.0
3
2
32
2
4
+
+
D
-2,3-Bis(3,5-di-t-butylsalicylideneamino)-1,4-buta-
([NiLCl] , 30), 411 ([NiL−H] , 65), 355.1 ([LH] ,
100).
nediol (13) was prepared from 4 and 3,5-di-t-butylsali-
cylaldehyde. Evaporation gave 13 as a yellow oil, m.p.
9
1
7
The complex of 10¦ with Cd(II) perchlorate was
1
6–103 °C (petroleum ether). H-NMR (CDCl ): l=
obtained as a colorless precipitate from MeOH-d solu-
4
3
1
4
tion. H-NMR (DMSO-d ): l=8.71, 8.64, 8.58 (s,
3.3 (brs, OH), 8.42 (s, NꢁCH), 7.37 (d, J=2 Hz, H ),
.07 (d, J=2.6 Hz, H ), 3.94 (dd, J=11.2, J=3.8
6
c
4
2
3
CHꢁN), 7.74, 7.20, 6.93 (m, Ar), 5.11, 5.08 (s, H ), 4.60,
2
e
2
3
4
.44 (m, ꢀCH CH ꢀ), 3.70, 3.51, 3.10, 2.94. FABMS;
Hz, H ), 3.87 (dd, J=11.2, J=6.6 Hz, H ), 3.65 (m,
H ), 1.42 (s, C(CH ) ), 1.27 (s, C(CH ) ). C-NMR
2
2
1
1’
+
13
m/z (relative intensity %): 567.0 ([CdL(ClO )] , 25),
4
2
3 3
3 3
+
+
4
67 ([CdL−H] , 7), 465 ([CdL−3H] , 15), 353.2
(
1
CDCl ): l=168.6 (d, NꢁCH), 158.2 (s, C ), 140.3,
3
b
+
(
[L−H] , 100).
36.9 (s, C ), 127.5, 126.4 (d, C ), 117.6 (s, C ), 72.2
d,f c,e a
The complex of 10¦ with Pb(II) perchlorate was
(
(
(
(
t, C ), 63.8 (d, C ), 35.1, 34.1 (s, C(CH ) ), 31.4, 29.4
1 2 3 3
obtained as a colorless precipitate from MeOH-d solu-
q, C(CH ) ). DCIMS; m/z: (relative intensity %): 553.4
[MH] 100); DEIMS; m/z (relative intensity %): 552.4
4
3
3
1
+
tion. m.p. 270 °C (dec.). H-NMR (DMSO-d ): l=
6
+
’
9.1–8.6 (m, NꢁCH), 7.80–7.10 (m, Ar), 4.48 (m,
[M ], 100), 434.3 (12), 276.2 ([M/2], 30). HR DEIMS;
+
ꢀCH CH ꢀ), 4.10 (m, H ), 3.70ꢀ3.40 (m, H ). FABMS;
2
2
1
+
2
m/z: 552.3927. Calc. for C H N O [M ] 552.3927.
34
52
−1
2
4
+
−
1
m/z (%):561 ([PbL−H] , 50), 559 ([PbL−3H] , 100),
55.1 ([LH] , 70).
UV–vis: umax (nm) (m, M cm ) (MeCN): 402 (101),
15 (7844), 256 (22 434), 215 (41 935); (MeOH): 403
1000), 315 (7500), 256 (23 000), 214 (40 500). [h] =
+
3
3
(
−
The complexation of 11¦ with nickel dichloride was
monitored by NMR up to a 1:1 ligand to metal ratio.
D
68° (c=2, CHCl3).
1
Final H-NMR spectrum (MeOH-d ): l=7.01 (m, Ar),
4
D
- 2,3 - Bis(3,5 - di - t - butylsalicylideneamino) - 1,4-
3
6
.74 (t, J=6.1, Ar), 4.93 (s, H ), 4.35, 4.07 (brd,
2
2
benzyloxybutane (11) was prepared from -2,3-di-
amino-1,4-benzyloxybutane (4a) and 3,5-di-t-butylsali-
D
J=6.5, ꢀCH CH ꢀ), 3.81 (brd, J=10, H4eq), 3.68
2
2
2
(
brd, J=10, H ), 2.72 (brs, H ). FABMS; m/z (rela-
4ax 9
cyl aldehyde. Evaporation gave a yellow oil product.
+
+
tive intensity %): 447.0 ([NiLCl] , 20), 355.1 ([LH] ,
00).
Salen)Mn(III) complexes were prepared according
to an existing procedure [16a].
-2,3-Bis(salicylideneamino)-1,4-butanediol
1
H-NMR (CDCl ): l=13.7 (brs, OH), 8.45 (s, NꢁCH),
3
1
4
7
.45 (d, J=2.4 Hz, H ), 7.34 (m, 5H ), 7.13 (d,
c
Ph
(
4
2
J=2.4 Hz, H ), 4.59 (d, J=12 Hz, CHHPh), 4.54 (d,
J=12 Hz, CHHPh), 3.90 (m, H ), 3.81 (dd, J=9.5,
J=5.2 Hz, H ), 3.63 (dd, J=9.5, J=6.5 Hz, H ),
.53 (s, C(CH ) ), 1.37 (s, C(CH3)3).
e
2
3
2
2
D
man-
2
3
1
1’
ganese(III) chloride (15) was prepared from 6 following
the original procedure [16a], but using only EtOH as
solvent, which was then evaporated and the product
13
1
C-NMR
3
3
(
1
CDCl ): l=167.9 (d, NꢁCH), 158.5, 140, 138, 136,
3
17.7 (s, Ar+Ph), 128.4, 127.7, 127.1, 126.2 (d, Ar+
was extracted by trituration with CHCl . FABMS; m/z
3
Ph), 73.3 (t, C ), 70.8 (t, CH Ph), 69.2 (d, C ), 35.1,
3
(
2
+
1
2
2
(
relative intensity %): 381 ([M−Cl] , 25), 380 ([(M−
4.1 (s, C(CH ) ), 31.5, 29.4 (q, C(CH ) ). CIMS; m/z
+
+
3
3
3 3
H)−Cl] , 95), 379 ([(M−2H)−Cl] , 100).
- 2,3 - Bis(salicylideneamino) - 1,4 - benzyloxybutane
+
relative intensity %): 732.7 ([MH] , 100), 366.4 ([M/2],
D
+
7 7 D 3
0), 91.1 ([C H ] , 55). [h] = +20° (c=4, CHCl ).
manganese(III) chloride (16) was similarly prepared
from 12. m.p. 275–278 °C. FABMS; m/z (relative in-
4
.3. Complexation of macrocycles 10¦and 11¦ with
+
tensity %): 561.1 ([M−Cl] , 100).
metal salts
D
-2,3-Bis(3,5-di-t-butylsalicylideneamino)-1,4-buta-
nediol manganese(III) chloride (17) was prepared from
All FABMS assignments were proven by comparison
with simulated isotope patterns.
1
3. m.p. 274 °C. FABMS; m/z (relative intensity %):
05.3 ([M−Cl] , 100). Anal. Found: C, 64.11; H, 8.16;
+
6
N, 4.45.Calc. for C H ClMnN O (MW 641.17): C,
34
50
2
4
1
4
.4. General procedure for H-NMR titration
6
3.69; H, 7.86; N, 4.37%.
-2,3-Bis(3,5-di-t-butylsalicylideneamino)-1,4-benzyl-
D
Metal salt solution (150 mmol in 0.5 ml) was added in
oxybutane manganese(III) chloride (18) was prepared
from 14. m.p. 265 °C. FABMS; m/z (relative intensity
portions of 10 ml (3 mmol) to the NMR tube containing
a solution of the macrocycle (30 mmol) in the deuter-
ated solvent (0.5 ml) and NMR spectra were taken
periodically.
+
%): 785.3 ([M−Cl] , 100). Anal. Found: C, 70.93; H,
7.93; Cl, 4.22; N, 3.55. Calc. for C H ClMnN O
48
62
2
4
(MW 821.42): C, 70.19; H, 7.61; Cl, 4.32; N, 3.41%.
Epoxidation of indene with (salen)Mn(III) complexes
was performed following a published procedure [16b],
but with certain changes. To a CH Cl (0.6 ml) solution
The complexation of 10¦ with nickel dichloride was
monitored by NMR up to a 1:1 ligand to metal ratio.
Final H-NMR spectrum (MeOH-d ): l=7.18 (t, J=
1
3
4
2
2

76
A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
of indene (0.6 g, 5.2 mmol) and catalyst (0.033 mmol),
bleach (0.7 M sodium hypochloride (10 ml) and sulfuric
acid (10%, 1 ml)) were added dropwise during 45 min at
Scholarship (to A.S.). Mr. Shimon Hauptman provided
valuable assistance with mass spectrometry and Mr.
Yuri Paskover contributed as an undergraduate re-
search student.
0
0
°C. The reaction is stirred for an additional 3.5 h at
°C. Work up [16b] gave epoxyindane as a yellow oil
and this was analysed by NMR with the chiral shift
reagent Eu(hfc)3.
References
Yields (ee): Catalyst 17: 0.34 g, 50% (48%); Catalyst
8: 0.35 g, 51% (30%).
1
[1] (a) New Supramolecular Host Systems. 14, taken from: A. Star,
PhD thesis, Tel-Aviv University, 1999;
(
(
(
b) A. Star, I. Goldberg, B. Fuchs, J. Phys. Org. Chem., in press;
c) A. Star, B. Fuchs, J. Org. Chem. 64 (1999) 1166;
d) A. Star, N.G. Lemcoff, I. Goldberg, B. Fuchs, Eur. J. Org.
4
.5. X-ray diffraction analysis of 6
The X-ray diffraction measurements were carried out
Chem. (1999) 2033;
e) A. Star, N.G. Lemcoff, I. Goldberg, B. Fuchs, Tetrahedron
on single crystals of 6 at the shown temperatures, on a
Nonius KappaCCD diffractometer, equipped with a
Cryosteam liquid nitrogen cooling device, and using
(
Lett. 38 (1997) 3573.
[2] (a) R.E. Valters, F. F u¨ l o¨ p, D. Korbonits, Adv. Heterocycle
Chem. (1996) 66;
Mo–K (u=0.7107 A
,
) radiation. Diffraction intensity
data were collected out to 2q=56.5° on 182 Phi-scan
1.0°) frames, and reduced to structure-factors by the
DENZO [17] and SCALEPACK [18] procedures. The crys-
tal structures were solved by direct methods (SHELXS
7) [19], and refined by full-matrix least-squares based
a
(
5
(
(
b) F. F u¨ l o¨ p, K. Pihlaja, J. Mattinen, G. Bern a´ th, J. Org. Chem.
2 (1987) 3821;
c) F. F u¨ l o¨ p, L. L a´ z a´ r, I. Pelczer, G. Bern a´ th, Tetrahedron 10
(
1988) 2993;
-
(d) A. Parkkinen, F. F u¨ l o¨ p, K. Pihlaja, Acta Chem. Scand. 44
(1990) 364.
9
2
[3] C. Hansch, A. Leo, R.W. Taft, Chem. Rev. 91 (1991) 165.
on F for all reflections (SHELXL-97) [20]. All non-hy-
drogen atoms were refined with anisotropic displace-
ment parameters.
[
4] A. Star, I. Goldberg, B. Fuchs, Angew. Chem. 112 (2000) 2797
Angew. Chem. Int. Ed. Engl. 39 (2000) 2685; a preliminary
(
communication of some of the results reported herein, in the
context of Dynamic Combinatorial Virtual Libraries).
C H N O , FW: 328.37, orthorhombic, space group
1
8
20
2
4
P2 2 2 .
[5] (a) E. Hadjoudis, Mol. Eng. 5 (1995) 301;
b) S.H. Alarc o´ n, A.C. Olivieri, M. Gonz a´ lez-Sierra, J. Chem.
Soc. Perkin Trans. 2 (1994) 1067;
c) S.H. Alarc o´ n, A.C. Olivieri, A. Nordon, R.K. Harris, J.
1
1 1
(
At 116 K: a=7.259(1), b=11.112(1), c=20.415(1)
3
−3
A, V=1646.7(3) A , Z=4, Dcalc=1.325 g cm , v
,
,
(
−
1
(Mo–K )=0.94 cm , R=0.052 for 2790 observa-
a
Chem. Soc. Perkin Trans. 2 (1996) 2293;
tions with F \4|(F ) and R=0.089 for all 3851
0
0
(d) F. Mansilla-Koblavi, J.A. Tenon, S. Toure, N. Ebby, J.
Lapasset, M. Carles, Acta Crystallogr. Sect. C 51 (1995) 1595.
[6] (a) T. Inabe, I. Luneau, T. Mitani, Y. Maruyama, S. Takeda,
Bull. Chem. Soc. Jpn. 67 (1994) 612;
−
3
unique data, S=1.00 and ꢀDzꢀ50.23 e A
,
.
At 295 K: a=7.362(1), b=11.120(1), c=20.709(1)
3
−3
A, V=1695.4(3) A , Z=4, Dcalc=1.286 g cm , v
, ,
−
1
(b) T. Inabe, New. J. Chem. 15 (1991) 129;
(
Mo–K )=0.92 cm , R=0.058 for 1839 observa-
a
(
c) T. Inabe, N. Hoshino, T. Mitani, Y. Maruyama, Bull. Chem.
Soc. Jpn. 62 (1989) 2245;
d) T. Inabe, N. Hoshino, T. Mitani, Y. Maruyama, Bull. Chem.
tions with F \4|(F ) and R=0.169 for all 4055
0
0
−
3
unique data, S=0.82 and ꢀDzꢀ50.15 e A
,
. All hy-
(
drogen atoms bound to O and N have been located
from difference-Fourier maps, but their positions were
not refined.
Soc. Jpn. 61 (1988) 4207.
[
[
7] (a) N. Brensciani Pahor, M. Calligaris, G. Nardin, L. Randaccio,
Acta Crystallogr. Sect. B 34 (1978) 1360;
(
b) N. Brensciani Pahor, M. Calligaris, P. Delise, G. Dodic, G.
Nardin, L. Randaccio, J. Chem. Soc. Dalton Trans. (1976) 2478.
8] (a) Y. Elerman, A. Elmali, M. Kabak, M. Aydin, M. Peder, J.
Chem. Cryst. 24 (1994) 603;
5. Supplementary material
(
b) J.P. Corden, W. Errington, P. Moore, M.G.H. Wallbridge,
Acta Crystallogr. Sect. C 53 (1997) 486;
c) J.C. Cannadine, J.P. Corden, W. Errington, P. Moore,
M.G.H. Wallbridge, Acta Crystallogr. Sect. C 52 (1996) 1014;
d) J.W. Yoon, T.-S. Yoon, W. Shin, Acta Crystallogr. Sect. C
3 (1997) 1685;
e) R. Kaitner, G. Pavlovi c´ , Acta Crystallogr. Sect. C 52 (1996)
2573.
Crystallographic data for the structural analysis have
(
been deposited with the Cambridge Crystallographic
Data Centre, CCDC nos. 149020 and 149021 for struc-
tures of 6. Copies of this information may be obtained
free of charge from The Director, CCDC, 12 Union
Road, Cambridge CB2 1EZ, UK (Fax: +44-1223-
(
5
(
336033; e-mail: deposit@ccdc.cam.ac.uk or www: http:/
[9] M. Yildiz, Z. Kilic, T. H o¨ kelek, J. Mol. Struct. 441 (1998) 1.
10] (a) L.G. Armstrong, L.F. Lindoy, Inorg. Chem. 14 (1975) 1322;
[
[
/
www.ccdc.cam.ac.uk).
(
1
(
b) L.F. Lindoy, H.C. Lip, L.F. Power, J.H. Rea, Inorg. Chem.
5 (1976) 1724;
c) S. Yamada, Coord. Chem. Rev. 190-192 (1999) 537.
Acknowledgements
11] (a) E.N. Jacobsen, in: I. Ojima (Ed.), Catalytic Asymmetric
Synthesis, VCH, New York, 1993, pp. 159–202;
This work was supported in part by a research grant
from the Israel Science Foundation and by an Intel
(
b) E.N. Jacobsen, in: L.A. Paquette (Ed.), Encyclopedia of
Reagents for Organic Synthesis, Wiley, Chichester, 1995;

A. Star et al. / Journal of Organometallic Chemistry 630 (2001) 67–77
c) M. Tokunaga, J.F. Larrow, F. Kakiuchi, E.N. Jacobsen, [16] (a) J.F. Larrow, E.N. Jacobsen, Org. Synth. 75 (1997) 1;
77
(
Science 277 (1997) 936.
12] (a) M.L. Wolfrom, F. Shafizadeh, J.O. Wehrm u¨ ller, R.K. Arm-
strong, J. Org. Chem. 23 (1958) 571;
(b) J.F. Larrow, E. Roberts, T.R. Verhoeven, K.M. Ryan, C.H.
Senanayake, E.N. Jacobsen, Org. Synth. 76 (1998) 46.
[17] Z. Otwinowski, DENZO: Program for Data Reduction and Inte-
gration, The University of Texas Southwestern Medical Center,
Dallas, TX, USA, 1985.
[
(
(
b) F.I. Carroll, J. Org. Chem. 31 (1966) 366;
c) P.W. Feit, O.T. Nielsen, J. Med. Chem. 10 (1967) 927.
[
[
13] J.-M. Lehn, Chem. Eur. J. 5 (1999) 2455.
14] Cambridge Structural Database System:
[18] Z. Otwinowski, SCALEPACK: Program for Merging and Scaling
of Intensities, The University of Texas Southwestern Medical
Center, Dallas, TX, USA, 1997.
[19] G.M. Sheldrick, SHELXS-86, in: G.M. Sheldrick, C. Kruger, R.
Goddard (Eds.), Crystallographic Computing 3, Oxford Univer-
sity Press, Oxford, 1985, pp. 175–189 (Acta Crystallogr. Sect. A,
46 (1990) 467–473).
[20] G.M. Sheldrick, SHELXL-97: Program for the Refinement of
Crystal Structures from Diffraction Data, University of G o¨ ttin-
gen, G o¨ ttingen, Germany, 1997.
(
(
(
(
a) F.H. Allen, O. Kennard, Chem. Des. Automation News 8
1993) 31;
b) F.H. Allen, O. Kennard, R. Taylor, Acc. Chem. Res. 16
1983) 146.
[
15] (a) W. Zhang, E.N. Jacobsen, J. Org. Chem. 56 (1991) 7;
b) M. Palucki, N.S. Finney, P.J. Pospisil, M.L. G u¨ ler, T. Ishida,
(
E.N. Jacobsen, J. Am. Chem. Soc. 120 (1998) 948 (and earlier
papers on the subject cited there).
.