Asymmetric cyclopropanation of olefins catalysed by Cu(i) complexes of chiral pyridine-containing macrocyclic ligands (Pc-L*)

Article abstract of DOI:10.1039/c2dt32347h

The synthesis and characterisation of copper(i) complexes of chiral pyridine-containing macrocyclic ligands (Pc-L*) and their use as catalysts in asymmetric cyclopropanation reactions are reported. All ligands and metal complexes were fully characterised, including crystal structures of some species determined by X-ray diffraction on single crystals. This allowed characterising the very different conformations of the macrocycles which could be induced by different substituents or by metal complexation. The strategy adopted for the ligand synthesis is very flexible allowing several structural modifications. A small library of macrocyclic ligands possessing the same donor properties but with either C₁ or C₂ symmetry was synthesized. Cyclopropane products with both aromatic and aliphatic olefins were obtained in good yields and enantiomeric excesses up to 99%.

Full text of DOI:10.1039/c2dt32347h

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Asymmetric cyclopropanation of olefins catalysed by
Cu(I) complexes of chiral pyridine-containing
macrocyclic ligands (Pc-L*)†
Cite this: Dalton Trans., 2013, 42, 2451
Brunilde Castano,a Stefano Guidone,a Emma Gallo,a Fabio Ragaini,a Nicola Casati,b
Piero Macchi,c Massimo Sistid and Alessandro Caselli*a
The synthesis and characterisation of copper(I) complexes of chiral pyridine-containing macrocyclic
ligands (Pc-L*) and their use as catalysts in asymmetric cyclopropanation reactions are reported.
All ligands and metal complexes were fully characterised, including crystal structures of some species
determined by X-ray diffraction on single crystals. This allowed characterising the very different confor-
mations of the macrocycles which could be induced by different substituents or by metal complexation.
The strategy adopted for the ligand synthesis is very flexible allowing several structural modifications.
A small library of macrocyclic ligands possessing the same donor properties but with either C1 or C2 sym-
metry was synthesized. Cyclopropane products with both aromatic and aliphatic olefins were obtained in
good yields and enantiomeric excesses up to 99%.
Received 4th October 2012,
Accepted 13th November 2012
DOI: 10.1039/c2dt32347h
www.rsc.org/dalton
containing the pyridine ring system with the aim of increasing
the stereochemical rigidity of the resulting complexes, which
Introduction
Tetraazamacrocycles, such as cyclen (1,4,7,10-tetraazacyclo- could result in an increase of their thermodynamic stability.2
dodecane) or its derivatives bearing functional groups on the The pyridine moiety may also provide a suitable site for a
nitrogen atoms, have attracted much attention in recent years further functionalization which may facilitate the binding
due to their ability to form stable complexes with a wide range ability of the ligand itself.
of metal ions. In this context, complexes derived from substi-
The construction of the macrocyclic ring system is the
tuted cyclen ligands and paramagnetic metal ions have been crucial step and it is dependent on the ring size.2 In the case
shown to be excellent probes for Magnetic Resonance Imaging of 12-membered ring systems, a 2 + 2 Richman–Atkins-type
(MRI).1
coupling3 was adopted and modified in order to improve the
The macrocyclic framework can be adjusted by changing chemical yields. By reacting 2,6-bis(chloromethyl)pyridine with
the nature of the donor atoms and the conformation flexibility 1,4,7-tritosyl-1,4,7-triazaheptane in anhydrous acetonitrile and
of the ring system in order to obtain pre-organized ligands in the presence of anhydrous potassium carbonate as hetero-
able to bind certain metal ions. In the field of MRI, some geneous base, conditions that approach a high dilution
years ago, we reported the synthesis of tetraazamacrocycles scenario, the yields of the cyclisation step were up to 94%
(Scheme 1). Indeed, under the Richman–Atkins protocol
(NaH/DMF as base/solvent system) the macrocycle A was isolated
in less than 30% yield.
aDipartimento di Chimica, Università di Milano, and ISTM-CNR, Via Golgi 19,
20133 Milano, Italy. E-mail: alessandro.caselli@unimi.it; Fax: (+39) 02 5031 4405;
Tel: (+39) 02 5031 4372
bLaboratory for Synchrotron Radiation – Condensed Matter, Swiss Light Source, Paul
Scherrer Institut, Villigen PSI CH-5232 Villigen, Switzerland
cDepartment of Chemistry and Biochemistry, University of Bern, Freiestrasse 3,
CH3012 Bern, Switzerland
dDip. Scienza e Alta Tecnologia, Università degli Studi dell’Insubria, Via Valleggio
11, 22100 Como, Italy
†Electronic supplementary information (ESI) available: Synthetic and analytical
data for all new compounds, spectra illustrating sample purity; crystallographic
tables and CIF files for compounds 4d, 4f and 5d. CCDC 904090, 904091,
904092. For ESI and crystallographic data in CIF or other electronic format see
DOI: 10.1039/c2dt32347h
Scheme 1 Synthesis of the macrocyclic ring system and numbering scheme
adopted.
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 2451–2462 | 2451

View Article Online
Paper
In order to extend our method to the synthesis of chiral
Dalton Transactions
Results and discussion
non racemic pyridine containing tetraazamacrocycles, we envi-
sioned that N-tosyl aziridine4 could be useful to prepare chiral
Design, synthesis and characterisation of complexes
4-substituted 1,4,7-triazaheptanes by nucleophilic ring The key step in the synthesis of the macrocyclic ligands, Pc-L*,
opening of 2 moles of N-tosyl aziridine with the appropriate is the assembly of the pyridine unit from the reaction of 2,6-
enantiomerically pure amine.5,6 The synthesized ligands were bis(chloromethyl)pyridine with an opportunely designed bis
used to prepare the corresponding paramagnetic metal ion (sulfonamide) under heterogeneous reaction conditions. The
complexes, after introduction of the appropriate pendant arms synthetic approach employed to obtain the starting bis(sulfo-
on the nitrogen atoms.5
namides) is reported in Scheme 2. There are two main steps
A few years ago, our group reported the synthesis and charac- involved in the synthesis. Firstly, ethanolamine or (L)-valinol
terisation of copper(I) complexes of new pyridine containing was reacted with 2 equivalents of tosyl chloride in the presence
tetraazamacrocyclic ligands7 and preliminary results on their of a weak base to afford aziridine derivatives 1a and 1b,6
use as catalysts in asymmetric cyclopropanation reactions.8 respectively. Use of triflic anhydride in the reaction with
These ligands combine the advantages of macrocyclic poly- valinol in dichloromethane afforded the aziridine 1c.19 Since
amines and those of aminopyridine ligands.9 Despite the evident the aziridines carry electron-withdrawing groups on the nitro-
potential of this class of C1 symmetric ligands, their application gen atom, they are expected to be susceptible to nucleophilic
in asymmetric catalysis has not been fully exploited.10 Indeed, attack. Thus, reaction with commercially available primary aryl
beside porphyrins and related ligands,11–15 tetraazamacrocycles amines in refluxing toluene resulted in the formation of the
have found a limited use in metal complexes catalytically active desired bis(sulfonamides) 3a–i in moderate to good yields
in enantioselective reactions. Recently Allen and co-workers (40–97%).
reported a new class of macrocyclic lanthanide complexes for the
enantioselective Mukaiyama aldol reactions.16,17
The nucleophilic attack of primary amines on 1-tosylaziri-
dine 1a occurs readily and use of two equivalents of aziridine
Transition metal-catalysed reactions of diazo compounds afforded bis(sulfonamides) 3a–3d directly, without isolation of
with alkenes have been widely used to prepare cyclopropanes. intermediate 2. When the chiral primary amines 1-benzylethyl-
In particular the copper-catalysed enantioselective version of amine and 1-naphthylethylamine, were employed both enan-
this reaction is now well established, and chiral C2 symmetric tiomers of the chiral sulfonamides 3c and 3d were synthesized
bidentate ligands such as bisoxazolines are the most widely and fully characterised. On the other hand, when bulkier (S)-2-
used.18 Here we report our findings concerning the asym- isopropyl-N-tosylaziridine was used, the ring opening reaction
metric cyclopropanation reaction of alkenes catalysed by was stopped after the addition of 1 equiv. of the aziridine and
copper(I) complexes derived from C1 and C2 symmetric pyri- the mono adducts 2e–i could be isolated, although in mode-
dine containing 12-membered tetraazamacrocyclic ligands rate yields (13–50%). It was interesting to note that in the
with two stereocenters on the ring skeleton and one stereo- present case, using naphthylalkylamines as nucleophiles, no
center on the carbon atom bonded to N-6 nitrogen atom. The large excess of amine was needed to obtain the mono adduct,
practical and efficient access to this class of compounds is rele- contrary to that previously reported in the case of the ring
vant for the development of new synthetic methodologies in opening of 2-isopropyl-N-(triflouromethylsulfonyl)aziridine
asymmetric catalysis.
with benzylamine.20 These reactions are extremely slow, but
Scheme 2 Synthesis of mono(sulphonamides) 2e–i and bis(sulphonamides) 3a–i.
2452 | Dalton Trans., 2013, 42, 2451–2462
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Scheme 3 Synthesis of the macrocyclic ligands (Pc-L*) 4a–i.
higher yields within shorter times were obtained by employing
microwave irradiation. It was found that after 3 h at 150 °C in
toluene, a 1.1 : 1 ratio of aziridine 1b and amine afforded sulfo-
namides 2 in 70% yield. Extending the reaction time and
using larger amounts of the aziridine relative to the amine did
not significantly improve the yield of bis(sulphonamide) 3.
Thus a two-step strategy was followed for their synthesis.
Firstly, the reaction was conducted at 150 °C under microwave
heating. Then, an excess of the aziridine was added to the reac-
tion mixture and the reaction was continued under conven-
tional heating. Under these conditions, bis(sulfonamides)
Scheme 4 Synthesis of the protonated salts 5a, 5c and 5d.
3e–g could be obtained in yields up to 74% after chromato- same stoichiometry and concentration of the reactants at
graphic purification.
130 °C (see ESI†), 20% of macrocycle 4i was observed after 2 h.
As expected, (S)-2-isopropyl-N-(triflouromethylsulfonyl)aziri- When the temperature was raised to 150 °C, 50% yield was
dine 1c is more reactive than the corresponding N-tosylaziri- obtained in just 3 h. Nevertheless, upon prolonged reaction
dine analogue and the reaction requires milder conditions.
The attack at the terminal position of the aziridine is highly observed to the reaction yield. Surprisingly, tosylsulfonamides
regioselective and we never observed competing ring 3a–g failed to yield the macrocycles under microwave heating.
times under these conditions, no beneficial effect was
a
opening reaction at the substituted carbon which was reported Further studies will be needed to clarify this aspect, in order to
in the case of 2-phenylaziridines.21 This methodology allows shorten the reaction times needed for the macrocyclisation
extensive structural modifications, since a large number of step and to improve the global yield. However, with the
primary amines can be used as nucleophiles in the ring employed methodology, a small library of macrocyclic ligands
opening reaction.
possessing different symmetries, steric and electronic proper-
Once bis(sulfonamides) 3a–i were obtained, the Richman– ties was easily obtained, starting from commercially available
Atkins cyclisation with 2,6-bis(chloromethyl)pyridine was products in a few synthetic steps. In order to investigate the
attempted. As already reported in the introduction, the macro- relative basicity of the four nitrogen atoms, we synthesized the
cyclisation was run under heterogeneous conditions.2 This protonated salts of some of the ligands, compounds 5a, 5c and
synthetic methodology avoids high dilution techniques and 5d (Scheme 4). The reaction was performed by treatment with
the 12 membered macrocycles 4a–i are obtained in 50–85% triflic acid in dichloroethane.
yields (Scheme 3).
Although the reaction was fast and quantitative as deter-
Although the employed conditions allowed the synthesis of mined by 1H NMR, the pure protonated salts were isolated
all the target macrocycles, in several cases (see ESI†), the time only in modest yields (ca. 50%). The products were fully
required was extremely long. In order to accelerate the reac- characterised, including X-ray structural determination in the
tion, we ran the synthesis in a microwave reactor. We initially case of the proton complex 5d (see below). The proton, as
focused on the synthesis of compound 4i, that under conven- expected for an ammonium salt, was located at high frequen-
1
tional heating required 110 h of reflux to yield the desired cies in the H NMR spectra (11.82 ppm for 5a, 11.24 ppm for
macrocycle in 40% yield after purification. Maintaining the 5c, 10.66 ppm for 5d in CDCl3 at room temperature). Worthy
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 2451–2462 | 2453

View Article Online
Paper
Dalton Transactions
of note is the fact that for compound 5d the 15N NMR chemical yellow solution. The addition of toluene to these solutions
shift for the pyridine nitrogen was observed at 282 ppm caused the precipitation of a white powder of complexes 6a or
(solvent CDCl3) instead of 312 ppm in the free ligand 4d, and 6c in moderate yields. Both complexes were characterised by
the resonance of N-6 (Scheme 1) was shifted from 39 to NMR spectroscopy. As previously reported by our group,7 the
52 ppm. This is consistent with a molecule where the proton is 1H NMR spectrum of complex 6a in CDCl3 displays a very low
coordinated in a bridging position between these two nitrogen symmetry. Two sets of signals are observed for the sulfona-
atoms (see Discussion of the X-ray data).
mide moieties and this can be due to the fact that only one of
Metal complex formation with ligands 4a–i was investigated the two nitrogen atoms bearing the tosyl substituent is truly
with different copper salts (Scheme 5). Treating ligands 4a and coordinated to the metal ion (Scheme 5). The same low sym-
4c with [Cu(OTf)]2·C7H8 gave yellow Cu(I) complexes that metry in solution is found also for complex 6c. The effect of
undergo oxidation quite readily. These complexes have been the copper complexation to the ligand was shown also by the
isolated and fully characterised. Both complexes are readily 15N NMR chemical shifts observed in CDCl3 solution. For
formed by treating a suspension of the copper salt in 1,2- instance, the pyridine signal that resonates at 312 ppm in free
dichloroethane (DCE) at room temperature, yielding a light ligand 4a, is located at 248 ppm in complex 6a. The nitrogen
bearing the benzyl substituent is shifted from 32 ppm in the
free ligand to 26 ppm upon complexation, whilst the less basic
N-tosyl substituted nitrogen atoms are affected only to a lower
extent, as expected.
Moreover, the presence of a naphthyl substituent seems to
confer a major stability to the copper(I) complexes and,
although the synthesis is better performed under a protecting
atmosphere, compounds 6b and 6d–i could be manipulated in
air without decomposition. The reason of this major stability
is due to the presence of the naphthyl group that can act as a
Scheme 5 Synthesis of the copper(I) complexes 6a–i.
further coordination site for the metal (Fig. 1). In these cases
Fig. 1 13C NMR spectra of complexes 6d (upper line) and 7d (lower line) in CDCl3.
2454 | Dalton Trans., 2013, 42, 2451–2462
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
also, low symmetry is retained in solution, as shown by NMR
studies. Again, the 15N NMR spectrum shows a marked shift
for the pyridinic nitrogen atom (from 313 to 245 ppm), while
the N-6 atom (Scheme 1), bonded to the stereogenic carbon, is
affected to a lower extent (from 39 to 51 ppm), similar to that
previously reported for the ammonium salt 5d. All these com-
plexes were isolated in good yields and fully characterised.
The η2 coordination mode of the naphthyl on copper(I) in
solution has been observed by 1H and 13C NMR studies in
CDCl3 solution. In the case of complex 6d, for example, a
marked low frequency shift of the naphthylic carbon a involved
in the η2 bond with copper from 124.5 ppm in the free ligand,
to 94.2 ppm in the complex (see Fig. 1 for labelling of the
carbon atoms a and b). Carbon b is affected to a lower extent
(low frequency shift from 124.6 to 118.2 ppm). A similar effect
1
can be observed also in the H NMR, with the proton directly
bound to carbon a shifted to higher frequencies (from 8.16 to
8.93 ppm). The observed coupling constant 1J (13C, 1H) of
149 Hz for carbon a provides hints of a partial re-hybridisation
state from sp2 to sp3. Carbon b, instead, is less affected and a
1J (13C, 1H) of 163 Hz is observed.22 The copper(I)–arene co-
ordinative interaction in a η2 unsymmetrical fashion has been
confirmed also by X-ray diffraction studies (vide infra) and the
Cu–C1 bond is significantly shorter than the Cu–C2 bond.
Although structurally characterised copper(I)–arene complexes
remain quite rare and, to the best of our knowledge this is the
first naphthyl complex reported, such a d–π interaction in
copper(I) complexes has been already systematically
investigated.23
In order to investigate the coordination behaviour of the
naphthyl group in the presence of an added ligand, we
decided to treat 6d with CO in dichloroethane solution
(Scheme 6). As revealed by IR in solution, a sharp band at
2111 cm−1 was present immediately after (5 min) exposure to
CO at atmospheric pressure. The observed frequency of coordi-
nated CO in complex 7d is higher than those reported in the
case of tetracoordinated copper(I) CO complexes of [bis(2-pyri-
dylmethyl)amine],24 probably due to the lower efficiency of ali-
phatic amines in transferring electron density to the metal
centre compared to pyridine.
The experiment was then repeated but exposing a CDCl3
solution of complex 6d in an NMR tube to 13C doped carbon
monoxide atmosphere. As reported in Fig. 1, a sharp signal at
171.3 ppm was observed, in a typical range for CO coordinated
to copper(I).25 On the other hand, signals relative to carbon a
and b were located at 122.9 and 127.0 ppm respectively, thus
clearly indicating that the naphthyl group is no longer coordi-
nated to the copper atom after the CO uptake. This result is of
particular relevance to catalysis, since it is reasonable to
assume that the naphthyl moiety can stabilise the coordi-
Scheme 6 Reaction of complex 6d with CO and CH3CN. Syntheses of com-
plexes 7–10d.
marked shift of all the signals, especially in the aliphatic
region. The 15N NMR spectrum shows that the position of the
pyridine nitrogen signal is unaffected while the sp3 nitrogen
atom bonded to the stereogenic carbon is located at 38 ppm.
The excess of acetonitrile can be easily removed under reduced
pressure to yield [Cu(CH3CN)4d]·(OTf), 8d, containing only
one coordinated acetonitrile molecule (νCN = 2250 cm−1), as
shown by integration of the CH3CN signal in the 1H NMR
spectrum. The same product, with different counter anions,
can also be synthesized by treating the free ligand 4d with
[Cu(CH3CN)4]·(PF6) or [Cu(CH3CN)4]·(BF4), to yield complexes
9d and 10d, respectively. Both complexes show the same
chemical shifts as 8d for the cation in the 1H NMR spectra,
indicating that the counter anion does not affect the symmetry
of the product in solution.
X-ray diffraction studies
nation sphere of the copper atom, but can be easily displaced The crystal structures of some of the species 4–10 have been
by incoming substrates. determined through X-ray diffraction on single crystal samples
The reactivity of complex 6d was also tested with aceto- (see Fig. 2–7 for the molecular geometries). This allowed char-
nitrile (Scheme 6). We first accomplished this by stirring a acterising the very different conformations of the cycles, which
dichloroethane solution of 6d in the presence of excess of could be induced by the substituents R, R1 and Ar or by the
acetonitrile (5 equivalents). The 1H NMR spectrum shows a complexation to the metal. In Table 1, we report a summary of
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 2451–2462 | 2455

View Article Online
Paper
Dalton Transactions
Fig. 2 Structure of compound 4a (thermal ellipsoids are shown at 50% prob-
ability level). To give an idea of the size available for coordination inside the
cycle, the distances N6⋯N12 and N3⋯N9 are 3.935(5) Å and 5.741(6) Å,
respectively.7.
Fig. 5 Structure of compound 5d (thermal ellipsoids are shown at 50% prob-
ability level; the [CF3SO3]− counter ion is omitted for clarity). Selected bond dis-
tances (Å) and angle (°): N6−H2 is fixed at 0.91, N6–H2–N12 153. For sake of
comparison with 4d, N6⋯N12 is 2.916(5) and N3⋯N9 5.090(5).
Fig. 3 Structure of compound 4d(13S) (thermal ellipsoids are shown at 50%
probability level). Distances within the coordination pocket are: N6⋯N12
4.122(5) Å and N3⋯N9 5.363(6) Å.
Fig. 6 Structure of compound 6d (thermal ellipsoids are shown at 50% prob-
ability level; the [CF3SO3]− counter ion is omitted for clarity). Selected bond dis-
tances (Å) and angle (°): Cu–N12 1.980(7), Cu–N6 2.151(6), Cu–N3 2.346(7),
Cu–N9 2.820(9), Cu–C1 2.017(8), Cu–C2 2.428(10), N3–Cu–N9 140.1(3).8
For sake of comparison with 4d and 5d, N6⋯N12 is 3.151(6) and N3⋯N9 is
4.902(6).
Fig. 4 Structure of compound 4f (thermal ellipsoids are shown at 50% prob-
ability level). Distances within the coordination pocket are: N6⋯N12 3.629(5) Å
and N3⋯N9 4.910(6) Å.
the conformations about each bond of the cycle. Apart from
the rigid conformations induced by the pyridine ring at the
12–1 and 11–12 bonds, all others show a fairly large flexibility,
so that only small similarities are found among the six species
studied in this work. In particular, it is obvious that complexa-
tion to Cu atom induces severe conformational changes to the
ring because the four nitrogen atoms must approach the metal
Fig. 7 Structure of compound 10d (thermal ellipsoids are shown at 50% prob-
ability level; the [CF3SO3]− counter ion is omitted for clarity). Selected bond dis-
tances (Å) and angle (°): Cu–N12 2.058(4), Cu–N6 2.111(5), Cu–N3 2.567(4),
Cu–N9 2.549(5), Cu–N(CH3CN) 1.898(5), N3–Cu–N9 139.1(1).8 For sake of com-
parison with 4d, 5d and 6d, N6⋯N12 is 3.306(5) and N3⋯N9 is 4.793(6).
In the free ligands, the conformation of the cycle brings the
for coordination (though, in 6d one nitrogen is not really tosyl bonded nitrogens antiparallel and the aromatic rings (the
bound to Cu, in spite of the favourable conformation, two tolyls and the phenyl or naphthyl) oriented far from each
see Fig. 6).
other. In 4f and 5d, one tosyl or the naphthyl is directed
2456 | Dalton Trans., 2013, 42, 2451–2462
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 1 Summary of the conformational features of the cyclic structures with (6 and 10) or without (4 and 5) Cu complexation, as determined from X-ray diffrac-
tion on single crystal samplesa
12–1
1–2
2–3
3–4
4–5
5–6
6–7
7–8
8–9
9–10
10–11
11–12
4a
4d
4f
5d
6d
10d
−ap
−ap
+ap
+ap
+ap
+ap
+sc
+sc
+sc
+sc
−sc
+sp
−sc
−sc
−ac
−ac
−sc
+sc
+ac
+ac
+sc
+ac
+ac
−ac
−ap
−ap
+sc
−sc
−sc
−sc
+sc
+ac
−ac
+ac
−sc
+ap
+sc
−ap
+ac
+ac
+sc
−sc
−sc
+sc
−ac
+sc
+sc
−sc
+ac
+sc
+sc
−ac
−ac
+sc
−sc
−ac
−sc
+sc
+sc
+sc
−sc
+ap
+ap
+ap
+ap
−ap
−ap
−sc
−sc
+ap
−ap
−sc
a Legend: sp = syn-periplanar; sc = syn-clinal; ac = anti-clinal; ap = anti-periplanar. +/− indicate the rotation.
Table 2 Cu(I)/4a catalysed cyclopropanation of α-methylstyrenea
Entry
Copper source
T
Solvent
Yieldb (%)
cis : transc
1
2
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C7H8
CuI
[Cu(OTf)]2·C6H6
[Cu(CH3CN)4]·BF4
Cu(OTf)2
r.t.
r.t.
r.t.
C2H4Cl2
C2H4Cl2
C2H4Cl2
C2H4Cl2
C2H4Cl2
C2H4Cl2
C2H4Cl2
C2H4Cl2
C2H4Cl2
C2H4Cl2
C2H4Cl2
CH2Cl2
31
33
32
89
45
89
10
16
32
43
88
44
10
50
47
19
37
42 : 58
42 : 58
42 : 58
43 : 57
43 : 57
43 : 57
44 : 56
45 : 55
44 : 56
45 : 55
43 : 57
64 : 36
50 : 50
77 : 23
72 : 28
66 : 34
40 : 60
3
4d
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
5d
6d
7d
CuCl2
8d
Cu(OAc)2·H2O
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C6H6
[Cu(OTf)]2·C6H6
9d,e
10d,f
11d,g
12d
13d
14d
15d
16d
17d
CH3CN
Toluene
Chlorobenzene
THF
h
C2H4Cl2
a Reactions were performed with equimolar amounts of Cu salt (3.0 × 10−2 mmol) and 4a in solvent. Cu–4a–EDA–α-mehylstyrene 1 : 1 : 35 : 170.
b Isolated yields based on EDA. c Determined by GC-MS. d Slow addition of EDA over 100 min. e In the presence of molecular sieves. f Cu–4a ratio
= 1 : 2. g Synthesized according to ref. 41. h Not distilled.
toward the centre of the cycle as it also occurs in the Cu Cyclopropanation reactions
complex 6d (whereas it is not possible in 10d, because of the
acetonitrile ligand). In 4f and 5d the bonds C2–N3 and C10–
N9 are associated with anti-clinal conformations, that makes
the pyridine ring lying on the average plane of the cycle,
EDA (EDA = ethyl diazoacetate). Catalytic reactions were run by
instead of being significantly tilted as in 4a and 4d and in the
adding EDA to a solution containing the olefin, the ligand and
Cu complexes. This conformation is less prompt for coordi-
nation to Cu, but likely a re-arrangement about C2–N3 and
C10–N9 is relatively easy.
The catalytic activity of copper complexes with ligand 4a in
cyclopropanation reactions was investigated. As a model reac-
tion we chose the cyclopropanation of α-methyl styrene with
the Cu salt (Cu–4a–EDA–olefin ratio 1 : 1 : 35 : 170), following
by IR spectroscopy the disappearance of the band due to the
stretching of the N2 moiety (ν = 2114 cm−1). We first tested the
In general, the intermolecular packing is not very tight, in
the absence of strong hydrogen bonding or other intermolecu-
ture and the modality of addition of the EDA solution. The
lar interactions. However, we cannot exclude that the observed
reaction by changing the copper salt, the solvent, the tempera-
results are summarised in Table 2. All reported yields were iso-
lated and based on EDA. In all cases cyclopropanes were
obtained as major products. However, the high volatility of the
products requires a careful isolation from the eluates. Fuma-
rate and maleate, the homo-coupling product of EDA, were the
only detected side products.
conformations could also be affected by packing effects. A full
investigation of the conformations of the ligands in isolation
is currently underway by means of theoretical gas phase
quantum chemical calculations.
The presence of S atoms, and the coordination to Cu,
allows a correct evaluation of the absolute configuration of the
species 4d, 4f, 5d, 6d and 10d, which is in agreement with
expectation.
As can be seen from the data reported in Table 2, the best
results in terms of yield were obtained at 0 °C by slow addition
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 2451–2462 | 2457

View Article Online
Paper
Dalton Transactions
of EDA with a syringe pump employing [Cu(OTf)]2·C6H6 as the of the reaction mixture. In all cases cyclopropanes were
copper source (the same results were obtained with the more obtained in good to excellent yields; again fumarate and
expensive toluene complex) in 1,2-dichloroethane (Table 2, maleate were the only detected side products and accounted
entry 4). Lower yields were observed when using copper(II) for the missing mass balance of the reactions. The presence of
sources such as CuCl2 or Cu(OAc)2 (Table 2, entries 7 and 8). the naphthyl group, although conferring a better stability to
Better results were obtained when employing copper(II) triflate the copper complex, does not alter the reaction course and
instead (Table 2, entry 6). It is reasonable to assume, anyway, ligand 4b essentially gave the same results compared to 4a
that copper must be reduced to oxidation state I by EDA under (Table 3, entry 1). In the case of chiral ligands, the presence of
the reaction conditions and that the active species is again a a naphthyl substituent seems to be beneficial and the enantio-
cationic copper(I) complex of the ligand having an outersphere selectivities obtained with ligand 4d are slightly better than
non coordinating triflate anion.26 In preliminary experiments, those with ligand 4c (Table 3, compare entries 8 and 9 with
freshly distilled solvents were employed. However, the use of entries 3 and 4). Further decreasing the temperature did not
molecular sieves during the reaction was found to be detri- improve the selectivity of the reaction (Table 3, entry 5). Identical
mental while yields were not severely affected by the use of not results were obtained whether preformed 6d or 4d/[Cu(OTf)]2·
distilled solvents (Table 2, entries 9 and 17, respectively). Good C6H6 were used (Table 3, entries 16 and 8). The presence of
yields were obtained in other chlorinated solvents such as acetonitrile in the reaction mixture slightly decreased the
dichloromethane and chlorobenzene and in toluene, while by yields, without affecting to a major extent the observed stereo-
using acetonitrile, lower yields were observed. In this case selectivities (Table 3, entries 6–7 and 17–19). This result
acetonitrile competes with the ligand and the substrates for suggests that the coordinated acetonitrile molecule dissociates
coordination to the copper atom (see later). A different expla- in solution before the activation of the EDA, and that the pre-
nation instead should be given for the low yield of cyclopro- catalyst is 4d whilst 6d, 7d and 8d are inactive. The inhibiting
panes formed when using THF. In this case the product effect was observed also upon addition of external acetonitrile
derived from the C–H insertion into the α-CH bond of the (Table 3, entry 20).
solvent was also observed (Table 2, entry 16).27
Best results in terms of enantioselectivities for both cis
We next studied the effect of the different ligands under the (88%) and trans (99%) isomers were obtained with the more
optimised conditions in C2H4Cl2 as solvent. The preformed sterically hindered ligand 4f (Table 3, entry 11). Ligand 4f has
complexes 6d, 8d, 9d and 10d were also tested. Results are a matching configuration of all the stereocenters and as a
reported in Table 3. All the reported yields in this case were result the observed enantioselection was significantly higher
determined by GC with the addition of 2,4-dinitrotoluene as than in the case of ligands 4d and 4e (Table 3, compare entries
1
internal standard, confirmed by quantitative H NMR analysis 8 and 10 with entry 13). On the other hand, a mismatching
Table 3 Cu(I)/Pc-L* complexes for asymmetric cyclopropanation of α-methylstyrenea
eed (%)
Entry
Pc-L*/Cu
T
Yieldb (%)
cis : transc
cis
trans
1
4b/[Cu(OTf)]2·C6H6
4c-(13S)/[Cu(OTf)]2·C6H6
4c-(13S)/[Cu(OTf)]2·C6H6
4c-(13R)/[Cu(OTf)]2·C6H6
4c-(13R)/[Cu(OTf)]2·C6H6
4c-(13R)/[Cu(CH3CN)4]·BF4
4d-(13R)/[Cu(CH3CN)4]·BF4
4d-(13R)/[Cu(OTf)]2·C6H6
4d-(13S)/[Cu(OTf)]2·C6H6
4e/[Cu(OTf)]2·C6H6
4f/[Cu(OTf)]2·C6H6
4f/Cu(OTf)2
4g/[Cu(OTf)]2·C6H6
4h/[Cu(OTf)]2·C6H6
4i/[Cu(OTf)]2·C6H6
6d
8d
9d
0 °C
r.t.
86
43 : 57
55 : 45
65 : 35
65 : 35
64 : 36
57 : 43
55 : 45
60 : 40
60 : 40
33 : 67
50 : 50
52 : 48
35 : 65
52 : 48
42 : 58
57 : 43
58 : 42
57 : 43
55 : 45
59 : 41
—
—
2e
3e
4
75(53)
90(70)
91(72)
80(59)
78(65)
70(56)
99(83)
99(83)
99(88)
98(57)
96
66
91
95
98(82)
80(72)
75(65)
85(73)
79(65)
−44
−50
50
53
33
45
53
−53
30
88
70
57
−21
23
55
49
−35
−38
38
38
36
44
65
−65
50
99
77
52
−16
59
66
63
0 °C
0 °C
−20 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
0 °C
5
6
7
8
9e
10
11
12
13
14e
15
16
17
18
19
20f
45
54
53
62
62
65
10d
4d-(13R)/[Cu(OTf)]2·C6H6
a Reactions were performed with equimolar amounts of Cu(I) salt (3.0 × 10−2 mmol) and Pc-L* with a S configuration in dichloroethane (5 mL).
Cu–4–EDA–α-mehylstyrene 1 : 1 : 35 : 170. b Yields based on EDA determined by GC (isolated yield). c Determined by GC. d Determined by chiral
HPLC; absolute configurations: cis-cyclopropanes were (1S,2R), trans-cyclopropanes were (1S,2S). e Absolute configurations: cis-cyclopropanes
were (1R,2S), trans-cyclopropanes were (1R,2R). f In the presence of added acetonitrile (5 equiv.).
2458 | Dalton Trans., 2013, 42, 2451–2462
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
Table 4 Asymmetric cyclopropanation of alkenes by Cu(I)/Pc-L*a
eed (%)
Entry
Pc-L*
Alkene
Yieldb (%)
cis : transc
cis
trans
1
2
4d-(13R)
4f
72(51)
97
50 : 50
38 : 62
33
99
50
87
3
4
4d-(13R)
4e
66(45)
79
53 : 47
25 : 75
34
33
45
40
5
6
7
4d-(13R)
4e
4f
88(68)
88
86
47 : 53
19 :81
42 : 58
36
42
66
50
53
78
8
9
10
4d-(13R)
4e
4f
64(56)
85
45
—
—
—
50
62
88
11e
4d-(13R)
81(54)
67 : 33
30
15
12
13
14
4d-(13R)
4e
4f
54(45)
55
74
>1 : 99
>1 : 99
>1 : 99
n.d.
n.d.
n.d.
57
28
76
15
4f
73
38 : 62
75
55
a Reactions were performed with slow addition of EDA over 100 min to a solution of [Cu(OTf)]2·C6H6 and PC-L* (3.0 × 10−2 mmol) in
dichloroethane (5 mL) at 0 °C. b Yields based on EDA determined by GC (isolated yield). c Determined by GC-MS. d Determined by chiral HPLC;
absolute configurations:28–30 for entries 1–7 cis-cyclopropanes were (1S,2R), trans-cyclopropanes were (1R,2R); entries 8–10: the absolute
configuration was (1S). For entries 11 and 15 the absolute configurations were not determined; for entries 12–14 trans-cyclopropane was
(1R,5R,6R). e Cu(I)–4d–EDA–olefin = 1 : 1 : 35 : 1750.
diastereoselection is observed: ligand 4d gave preferentially
Under optimal conditions, other alkenes were employed to
the cis isomer, while ligand 4e the trans and, in the case of determine the substrate scope of the cyclopropanation reaction
ligand 4f, equal amounts of both isomers were obtained. catalysed by the Pc-L*/[Cu(OTf)]2·C6H6 system in dichloro-
Noticeably, the reaction with Cu(OTf)2 still gave very good ethane (Table 4). At a Cu(I)–EDA–alkene ratio of 1 : 35 : 170, the
yields, but a significant drop in the enantioselection (Table 3, formed complex catalysed the reaction of all the tested sub-
entry 12).
strates yielding the cyclopropanes in good yields and enantio-
The two diastereoisomers 4g and 4f afforded the same selectivities. The absence of α-substituents on the styrene
enantiomer albeit in different enantiomeric excess (Table 3, affected the diastereoselectivity of the reaction (Table 4, entries
entry 13). The experimental results show that better enantio- 1–7) yielding preferentially to the trans isomer (up to a 4 fold
meric excesses were obtained if a stereogenic carbon is present excess in the case of 4-Cl-styrene, entry 6). Steric hindrance at
in position 13 (Scheme 3). On the other hand, the diastereo- the α position does not hamper the reaction and good yields
selection in favour of the cis-isomer is influenced by the pres- were obtained with 1,1-diphenyl ethylene (Table 4, entries
ence of bulky substituents on the macrocycle. The presence of 8–10).
the i-propyl moieties, in fact, confine the macrocycle in a more
rigid environment (Fig. 4).
The catalytic activity is not confined to styrenic double
bonds. With 2,5-dimethyl-2,4-hexadiene, an important precur-
Surprisingly, when employing ligand 4h we obtained the sor to the chrysanthemic acid synthesis,32 the catalytic reaction
opposite enantiomer (Table 3, entry 14). The stereocenters of yielded the desired cyclopropanes (cyclopropanation of only
this molecule are exactly in the same configuration as those of one double bond was observed)33 and with reasonable diaster-
4e. It is reasonable to assume that trifyl moieties induces a eoselectivities, if a large excess of the olefin is employed
different conformation of the macrocyclic ring system.
(Table 4, entry 11).
t
When employing the bulkier tert-butyl diazoacetate, BuDA,
Good yields (attack only at the non-substituted double
(condition as in the caption to Table 2, 4d/[Cu(OTf)]2·C6H6 as bond) and excellent diastereoselectivities in favour of the trans
catalyst) both lower yields (28%) and a decrease of enantio- isomer (>99%) were obtained with 2-methylfuroate, (Table 4,
selectivity (ee cis 12%, trans 46%) were observed with an inver- entries 12–14).34 From this precursor, some years ago Reiser
sion of the diastereoselectivity (cis : trans = 27 : 73). Similar reported a very elegant synthesis of roccellaric acid.35
effects on increasing the size of the diazoacetate are not unpre-
Non-activated aliphatic alkenes are known to be less reac-
cedented and have been explained in terms of overcrowding of tive in comparison with aromatic or activated alkenes towards
the transition state.31
copper mediated cyclopropanations.36 Chiral catalysts for the
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 2451–2462 | 2459

View Article Online
Paper
Dalton Transactions
enantioselective cyclopropanation of aliphatic 1-alkenes are triflate benzene complex, copper(I) tetrakis–acetonitrile esa-
very often trans selective,37,38 although more recent examples fluorophosphate complex and copper(I) tetrakis–acetonitrile
of highly cis selective catalysts have been reported.39,40 When tetrafluoroborate complex were synthesized following literature
1-octene was reacted with EDA in the presence of Cu(I)/4f the methods.41 All other starting materials were commercial pro-
corresponding cyclopropane derivative was obtained in 73% ducts and were used as received.
yield, thus extending the scope of Pc-L* copper(I) complexes in
asymmetric synthesis (Table 4, entry 15).
X-ray structure determination
Data collections for the crystal structure determinations were
carried out using a Bruker Apex II difractometer (4a, 5d, 6d
and 10d) or an Agilent SuperNova Mo microsource, Al fil-
Conclusions
tered42 and working at 50 kV and 0.8 mA (compounds 4d and
In conclusion, we have prepared and fully characterised a new 4f). All crystal structures were solved by direct methods using
family of Cu(I) complexes with chiral pyridine containing SHELXS97 and refined with SHELX97,43 within the WINGX
macrocyclic ligands (Pc-L*). The strategy adopted for the syn- suite of programs.44 H atoms were rigidly modelled on the
thesis of these macrocyclic ligands is very flexible and allows riding C or N atoms.
for structural modification. A small library of ligands posses-
Details of the crystal structures and data collections of 4a,7
sing the same donor properties but with either C1 or C2 sym- 6d and 10d8 have been already reported. Crystallographic
metry was synthesized. Their use in catalytic asymmetric information files and Tables with crystallographic data for 4d,
cyclopropanation reactions was investigated. Our group is cur- 4f and 5d are reported as ESI.† Here we report only the major
rently investigating the nature of the active intermediates and features of each crystal structure.
the utility of pyridine containing macrocyclic ligands in other
4a. Crystals suitable for an X-ray structural determination
asymmetric reactions. The structural features of these ligands were obtained by crystallization of 4a from a dichloromethane/
allow further modifications in order to modulate the proper- toluene solution (Fig. 2). In the absence of chirality, this com-
ˉ
ties of the corresponding complexes. Furthermore, since mono pound crystallize in centrosymmetric P1.
adducts 2e–i could be isolated in good yields and purity, the
4d(13S). Crystals suitable for an X-ray structural determi-
synthesis of unsymmetrical bis(sulfonamides) can be envi- nation were obtained by crystallization from a ethyl acetate/
saged and this will be the subject of forthcoming work from n-hexane solution (Fig. 3). The species crystallizes in chiral
our group.
P212121. As reported in Table 1, the conformation of 4d is not
so dissimilar from 4a, albeit not identical. The absolute con-
figuration is confirmed (Flack parameter −0.07(10)).
4f(13R4S8S). Crystals suitable for X-ray diffraction were
obtained from CDCl3. The samples were stable in air, and a
data collection was run at ambient conditions. The compound
Experimental section
General information
NMR spectra were recorded on Bruker Avance 300-DRX or crystallizes in P212121 space group, which is as expected a
Avance 400-DRX spectrometers. Chemical shifts (ppm) are chiral space group (see Fig. 4). The presence of S atoms allows
reported relative to TMS. The 1H NMR signals of the com- the determination of the absolute configuration, with
pounds described in the following have been attributed by sufficient accuracy (Flack parameter 0.01(6)).
COSY and NOESY techniques. Assignments of the resonance
5d(13R). Crystals suitable for an X-ray structural determi-
in 13C NMR were made using the APT pulse sequence and nation were obtained by crystallization of 5d from a dichloro-
HSQC and HMBC techniques. The 15N NMR signals of the ethane solution (Fig. 5). The species crystallizes as a salt of
compound described have been attributed by HMBC tech- [CF3SO3]− in chiral P212121. As described above, 5d and 4f have
nique. Infrared spectra were recorded on a BIO-RAD FTS-7 some similarities in the conformation of the cycle, but overall
spectrophotometer. Elemental analyses and mass spectra were they are quite different (see Table 1), mainly because of the
recorded in the analytical laboratories of Milan University. intra-molecular N–H⋯N hydrogen bond (dN–N 2.916 Å) occur-
GC-MS analysis were performed on a Shimadzu GCMS- ring between the protonated N6 and the pyridine N12. The
QP5050A instrument. Optical rotation were measured on a larger perturbation at N6 is in keeping with the assigned posi-
Perkin Elmer instruments model 343 plus; [α]D values are tion of the H atom, which was also tentatively refined (result-
given in 10−1 ° cm2 g−1. Microwave assisted reactions were ing in dN6–H6 0.7 Å). As this refinement was somewhat
performed with a MILESTONE® microSYNT multimode labsta- unstable, in the final model it was constrained into the stereo-
tion, using 12 mL sealed glass vessels. The internal tempera- chemically predicted position. The absolute configuration is
ture was detected with a fiber optic sensor. Unless otherwise confirmed by the X-ray analysis (Flack parameter −0.02(10)).
specified, all the reactions were carried out in a dinitrogen
atmosphere employing standard Schlenk techniques and mag- nation were obtained by layering with benzene a dichloro-
netic stirring. Solvents were dried prior to use by standard pro- ethane solution from ethyl acetate/n-hexane solution
6d(13S). Crystals suitable for an X-ray structural determi-
a
cedures and stored under dinitrogen. α-Methylstyrene was (Fig. 6). The space group is P212121 and the absolute configur-
distilled over CaH2 and stored under dinitrogen. Copper(I) ation could be confirmed with sufficient precision (Flack
2460 | Dalton Trans., 2013, 42, 2451–2462
This journal is © The Royal Society of Chemistry 2013

View Article Online
Dalton Transactions
Paper
parameter 0.04(3)). The main feature of this complex is the n-hexane–ethyl acetate = 6 : 4 as eluent, obtaining a white solid
1
loose coordination of N9 to the Cu complex (2.820(9) Å), which 4f-(13R,4S,8S) (0.471 g, 0.626 mmol, 57%). H NMR (300 MHz;
implies a tetra-coordination, instead of a penta-coordination.
CDCl3; T = 300 K) δ 8.58 (1H, d, J = 8.1 Hz, ArH), 7.89 (1H, m,
10d(13S). Crystals suitable for an X-ray structural determi- ArH), 7.83 (1H, d, J = 8.1 Hz, ArH), 7.69 (1H, d, J = 8.1 Hz, ArH),
nation were obtained by crystallisation from dichloroethane/ 7.64–7.42 (8H, m, ArH), 7.12 (4H, d, J = 7.8 Hz, ArH), 7.05 (2H,
n-hexane. The structure of complex 10d, (Fig. 7) shows the m, ArH), 5.18 (1H, m, H13), 4.72 (2H, m, H2), 3.94–3.82 (6H, m,
copper atom placed in the middle of the macrocyclic cavity, H), 2.63–2.60 (2H, m, H), 2.40 (2H, m, H) overlapping with 2.33
again in a fivefold coordination site produced by the four (6H, s, CH3), 1.57 (3H, d, J = 6.5 Hz, CH3), 0.44 (12H, m, CH3).
ligand nitrogens and the acetonitrile. At variance from 6d, in 13C NMR (75 MHz; CDCl3; T = 300 K) δ 156.5 (C), 142.5 (C),
10d Cu is therefore truly pentacoordinated in a strongly dis- 142.2 (C), 138.9 (C), 137.0 (CH), 134.2 (C), 131.7 (C), 129.2
torted trigonal bipyramidal geometry. N3 and N9 occupy par- (CH), 128.9 (CH), 127.8 (CH), 127.0 (CH), 126.0 (CH), 125.9
ticularly elongated axial sites (Cu–N ∼ 2.5 Å), whereas N6, N12 (CH), 125.8 (CH), 125.2 (CH), 124.0 (CH), 120.7 (CH), 66.5
and the acetonitrile are in equatorial positions. As expected, (CH), 50.8 (CH2), 49.5 (CH2), 30.0 (CH), 23.8 (CH3), 21.5 (CH3),
the naphthyl group has been displaced from the coordination 21.1 (CH3). A signal relative to a CH was not detected.
sphere of the metal by the incoming acetonitrile molecule. 15N NMR (40 MHz; CDCl3; T = 300 K) δ 33.2 (NC). The signal
The complex crystallizes in P21, and the absolute configuration relative to N-Ts and N12 were not detected. Elemental Analysis:
is well established (Flack parameter −0.003(16)).
Synthesis of 1,7-ditosyl-2,6-[(S,S)-iso-propyl]-4-[(R)-1- 68.6; H, 7.0; N, 7.4%. m/z 753 (M + 1).
naphthylethyl]-1,4,7-triazaheptane, 3f-(1R,2S,2′S). A solution Synthesis of 6f. Copper(I) triflate benzene complex
Found: C, 68.4; H, 7.4; N, 7.1%. Calc. for C43H52N4O4S2: C,
of (S)-2-isopropyl-1-tosylaziridine 1b (0.781 g, 3.21 mmol) and (0.0298 g, 0.0591 mmol) was added to a solution of macrocycle
(R)-1-(1-naphthyl)ethyl amine (0.514 g, 3.0 mmol) in toluene 4f-(13R,4S,8S) (0.0891 g, 0.118 mmol) in dichloroethane
(21 mL) was added in a microwave tube. The solution was (10 mL). The solution was stirred at room temperature for one
stirred and heated by microwave irradiation for 3 h at 150 °C. hour, concentrated to 5 mL and then 10 mL of n-hexane were
The resulting mixture, without any purification, was added to layered. Then the solid was filtered and dried in vacuo under
a
solution of (S)-2-isopropyl-1-tosylaziridine 1b (0.785 g, nitrogen, obtaining complex 6f as a white solid (0.111 g,
3.28 mmol) in distilled toluene (4 mL) and was stirred and 0.115 mmol, 97%). 1H NMR (300 MHz; CDCl3; T = 300 K) δ
heated under reflux for 45 h. The mixture was dried and puri- 8.68 (1H, d, J = 8.7 Hz, ArH), 7.93 (1H, d, J = 8.0 Hz, ArH), 7.87
fied by silica gel chromatography using n-hexane–ethyl acetate (4H, d, J = 8.1 Hz, ArH), 7.80–7.69 (3H, m, ArH), 7.62–7.40 (7H,
= 7 : 3 as eluent, obtaining a light yellow solid 3f-(1R,2S,2′S) m, ArH), 7.23–7.16 (2H, m, ArH), 5.95 (1H, q, J = 6.6 Hz, CH),
(0.785 g, 1.21 mmol, 40%). 1H NMR (400 MHz; CDCl3; T = 5.25 (1H, d, J = 20 Hz, CH2), 4.79 (1H, d, J = 12.7 Hz, CH2), 4.71
300 K) δ 7.94 (1H, m, ArH) 7.79 (4H, d, J = 8.0 Hz, ArH) overlap- (1H, d, J = 14.6 Hz, CH2), 4.62 (1H, d, J = 20 Hz, CH2), overlap-
ping with 7.79 (1H, m, ArH), 7.73 (1H, dd, J = 5.8 Hz, J = ping with 4.61 (1H, m, CH), 4.01 (1H, d, J = 14.6 Hz, CH2), 2.78
3.4 Hz, ArH), 7.45–7.42 (2H, m, ArH), 7.40–7.39 (2H, m, ArH), (1H, d, J = 12.7 Hz, CH2), 2.57 (3H, s, CH3), overlapping with
7.29 (4H, d, J = 8.0 Hz, ArH), 4.66 (2H, d, J = 7.9 Hz, NH), 4.59 2.57 (1H, m, CH), 2.52 (3H, s, CH3), 2.42–2.37 (1H, m, CH2),
(1H, q, J = 6.6 Hz, CH), 3.26 (1H, m, CH), 2.46 (2H, dd, J = 13.4 2.30–2.15 (2H, m, CH2 and CH), 2.10 (3H, d, J = 6.8 Hz, CH3),
Hz, J = 5.1 Hz, CH2), 2.41 (6H, s, CH3), 1.55 (2H, m, CH), 1.20 1.59 (1H, m, CH), 0.90 (3H, d, J = 6.7 Hz, CH3), 0.68 (3H, d, J =
(3H, d, J = 6.6 Hz, CH3) 0.56 (6H, d, J = 6.9 Hz, CH3), 0.19 (6H, 6.2 Hz, CH3), 0.29 (3H, d, J = 6.5 Hz, CH3), −0.49 (3H, d, J =
d, J = 6.8 Hz, CH3). 13C NMR (75 MHz; CDCl3; T = 300 K) δ 6.2 Hz, CH3). 13C NMR (75 MHz; CDCl3; T = 300 K) δ 155.7 (C),
143.4 (C), 138.2 (C), 138.0 (C), 134.1 (C), 132.1 (C), 129.7 (CH), 151.2 (C), 145.6 (C), 145.3 (C), 140.4 (CH), 134.8 (C), 134.5 (C),
128.8 (CH), 128.2 (CH), 127.3 (CH), 125.6 (CH), 125.5 (CH), 133.9 (C), 132.1 (C), 131.5 (CH), 130.4 (C), 129.8 (CH), 129.3
125.1 (CH), 125.0 (CH), 123.9 (CH), 56.6 (CH), 53.6 (CH), 52.3 (CH), 128.9 (CH), 127.7 (CH), 127.3 (CH), 126.4 (CH), 125.2
(CH2), 27.6 (CH), 21.6 (CH3), 19.3 (CH3), 14.8 (CH3), 12.4 (CH), 124.8 (CH), 124.5 (CH), 123.9 (CH), 122.8 (C), 64.7 (CH),
(CH3). 15N NMR (40 MHz; CDCl3; T = 300 K) δ 94.0 (NHTs), 62.0 (CH), 57.4 (CH2), 57.1 (CH2), 56.4 (CH), 55.5 (CH2), 46.8
38.3 (NHC). The signal relative to N-Ts was not detected. (CH2), 29.9 (CH), 27.1 (CH), 24.5 (CH3), 22.4 (CH3), 21.9 (CH3),
Elemental Analysis: Found: C, 66.6; H, 7.6; N, 6.5%. Calc. for 21.3 (CH3), 20.3 (CH3), 18.5 (CH3). 19F NMR (282 MHz; CDCl3;
C36H47N3O4S2: C, 66.5; H, 7.3; N, 6.5%. [α]2D0 = −47.22 (c 1.37 in T = 300 K) δ −78.58 (s). Elemental Analysis: Found: C, 54.7; H,
CHCl3).
5.2; N, 5.7%. Calc. for C44H52CuF3N4O7S3: C, 54.7; H, 5.4; N,
Synthesis of 6-[(R)-1-naphthylethyl]-3,9-ditosyl-3,8-[(S,S)-iso- 5.8%.
propyl]-3,6,9,15-tetraazabicyclo[9,3,1]pentadeca-1(15),11,13-triene,
4f-(13R,4S,8S). A solution of amine 3f-(1R,2S,2′S) (0.715 g,
General procedure for the catalytic cyclopropanation reactions
1.10
mmol),
2,6-bis(chloromethyl)pyridine
(0.194
g, In
a
typical experiment, [Cu(OTf)]2·(C6H6) (0.0075 g,
1.10 mmol) and micronized anhydrous potassium carbonate 0.015 mmol), the ligand (0.020 g (4d), 0.030 mmol) and
(0.608 g, 4.40 mmol) in distilled acetonitrile (37 mL) was α-methyl styrene (0.650 mL, 5.0 mmol) were dissolved in dis-
stirred and heated under reflux for 116 h. The resulting tilled dichloroethane (5 mL) and the solution stirred for one
mixture was washed with water end extracted with ethyl hour at 0 °C. Then a dichloroethane solution (1 mL) of EDA
acetate, dried and purified by silica gel chromatography using (0.114 g, 0.105 mL, 1 mmol) was slowly added by a syringe
This journal is © The Royal Society of Chemistry 2013
Dalton Trans., 2013, 42, 2451–2462 | 2461

View Article Online
Paper
Dalton Transactions
pump over 100 min. The reaction was monitored by IR, follow- 17 Y. Mei, P. Dissanayake and M. J. Allen, J. Am. Chem. Soc.,
ing the disappearance of the band due to the stretching of N2
2010, 132, 12871–12873.
moiety at 2114 cm−1. The reaction was considered to be 18 H. Pellissier, Tetrahedron, 2008, 64, 7041–7095.
finished when the absorbance of the EDA was below 0.03 (by 19 M. Cernerud, H. Adolfsson and C. Moberg, Tetrahedron:
using a 0.1 mm thick cell). 2,4-Dinitrotoluene was added as
Asymmetry, 1997, 8, 2655–2662.
internal standard and the solution was analysed by GC. The 20 F. Lake and C. Moberg, Eur. J. Org. Chem., 2002, 3179–3188.
solution was then evaporated to dryness in vacuo and analysed 21 P. Sudhakar and G. Sundararajan, J. Polym. Sci., Part A:
1
by H NMR. The residue purified by silica gel chromatography
Polym. Chem., 2006, 44, 4006–4014.
(eluent ethyl acetate–n-hexane = 0.3 : 10).
22 C. Martín, J. M. a. Muñoz-Molina, A. Locati, E. Alvarez,
F. Maseras, T. s. R. Belderrain and P. J. Pérez, Organometal-
lics, 2010, 29, 3481–3489.
23 T. Osako, Y. Tachi, M. Doe, M. Shiro, K. Ohkubo,
S. Fukuzumi and S. Itoh, Chem.–Eur. J., 2004, 10, 237–246.
24 H. R. Lucas, G. J. Meyer and K. D. Karlin, J. Am. Chem. Soc.,
2010, 132, 12927–12940.
Acknowledgements
The financial support of MiUR is gratefully acknowledged.
25 M. Kujime, T. Kurahashi, M. Tomura and H. Fujii, Inorg.
Chem., 2007, 46, 541–551.
26 R. G. Salomon and J. K. Kochi, J. Am. Chem. Soc., 1973, 95,
3300–3310.
Notes and references
1 E. Terreno, D. D. Castelli, A. Viale and S. Aime, Chem. Rev., 27 H. M. L. Davies and R. E. J. Beckwith, Chem. Rev., 2003,
2010, 110, 3019–3042. 103, 2861–2903.
2 S. Aime, M. Botta, S. G. Crich, G. B. Giovenzana, G. Jommi, 28 A. Berkessel, P. Kaiser and J. Lex, Chem.–Eur. J., 2003, 9,
R. Pagliarin and M. Sisti, Inorg. Chem., 1997, 36,
2992–3000.
3 J. E. Richman and T. J. Atkins, J. Am. Chem. Soc., 1974, 96,
2268–2270.
4746–4756.
29 T. Niimi, T. Uchida, R. Irie and T. Katsuki, Adv. Synth.
Catal., 2001, 343, 79–88.
30 A. Nakamura, A. Konishi, R. Tsujitani, M. Kudo and
S. Otsuka, J. Am. Chem. Soc., 1978, 100, 3449–3461.
4 S. Aime, M. Botta, L. Frullano, S. G. Crich, G. Giovenzana,
R. Pagliarin, G. Palmisano, F. R. Sirtori and M. Sisti, J. Med. 31 A. M. Harm, J. G. Knight and G. Stemp, Tetrahedron Lett.,
Chem., 2000, 43, 4017–4024. 1996, 37, 6189–6192.
5 S. Aime, E. Gianolio, D. Corpillo, C. Cavallotti, 32 M. Itagaki and K. Suenobu, Org. Process Res. Dev., 2007, 11,
G. Palmisano, M. Sisti, G. B. Giovenzana and R. Pagliarin,
Helv. Chim. Acta, 2003, 86, 615–632.
6 B. M. Kim, S. M. So and H. J. Choi, Org. Lett., 2002, 4,
949–952.
7 B. Castano, T. Pedrazzini, M. Sisti, E. Gallo, F. Ragaini,
N. Casati and A. Caselli, Appl. Organomet. Chem., 2011, 25,
824–829.
509–518.
33 A. G. M. Barrett, D. C. Braddock, I. Lenoir and H. Tone,
J. Org. Chem., 2001, 66, 8260–8263.
34 C. Böhm, M. Schinnerl, C. Bubert, M. Zabel, T. Labahn,
E. Parisini and O. Reiser, Eur. J. Org. Chem., 2000,
2955–2965.
35 C. Bohm and O. Reiser, Org. Lett., 2001, 3, 1315–1318.
8 A. Caselli, F. Cesana, E. Gallo, N. Casati, P. Macchi, 36 N. S. Youssef, A. M. A. El-Seidy, M. Schiavoni, B. Castano,
M. Sisti, G. Celentano and S. Cenini, Dalton Trans., 2008,
4202–4205.
F. Ragaini, E. Gallo and A. Caselli, J. Organomet. Chem.,
2012, 714, 94–103.
9 W. Ye, D. M. Ho, S. Friedle, T. D. Palluccio and E. V. Rybak- 37 J. A. Miller, W. Jin and S. T. Nguyen, Angew. Chem., Int. Ed.,
Akimova, Inorg. Chem., 2012, 51, 5006–5021. 2002, 41, 2953–2956.
10 S. J. Archibald, Annu. Rep. Prog. Chem., Sect. A, 2009, 105, 38 K. Ito and T. Katsuki, Tetrahedron Lett., 1993, 34,
297–322. 2661–2664.
11 D. Intrieri, A. Caselli and E. Gallo, Eur. J. Inorg. Chem., 39 S. F. Zhu, X. Xu, J. A. Perman and X. P. Zhang, J. Am. Chem.
2011, 5071–5081. Soc., 2010, 132, 12796–12799.
12 C.-M. Che and J.-S. Huang, Chem. Commun., 2009, 40 C. Bonaccorsi and A. Mezzetti, Organometallics, 2005, 24,
3996–4015. 4953–4960.
13 S. Zhu, X. Cui and X. P. Zhang, Eur. J. Inorg. Chem., 2012, 41 K. M. Gillespie, C. J. Sanders, P. O’Shaughnessy,
430–434.
I. Westmoreland, C. P. Thickitt and P. Scott, J. Org. Chem.,
2002, 67, 3450–3458.
14 M. P. Doyle, Angew. Chem., Int. Ed., 2009, 48, 850–852.
15 J. C. Marchon and R. Ramasseul, in The Porphyrin Hand- 42 P. Macchi, H.-B. Buergi, A. S. Chimpri, J. Hauser and
book, ed. K. M. Kadish, K. M. Smith and R. Guilard,
Academic Press, San Diego, CA, 2003, vol. 11, pp. 75–132.
16 Y. Mei, D. J. Averill and M. J. Allen, J. Org. Chem., 2012, 77,
5624–5632.
Z. Gal, J. Appl. Crystallogr., 2011, 44, 763–771.
43 G. M. Sheldrick, Acta Crystallogr., Sect. A: Fundam. Crystal-
logr., 2008, 64, 112–122.
44 L. J. Farrugia, J. Appl. Crystallogr., 1999, 32, 837–838.
2462 | Dalton Trans., 2013, 42, 2451–2462
This journal is © The Royal Society of Chemistry 2013